From b7df4531df9ed737bfd2d3aef1e322979cc53236 Mon Sep 17 00:00:00 2001
From: Fize Jacques <jacques.fize@cirad.fr>
Date: Thu, 3 Dec 2020 09:38:26 +0100
Subject: [PATCH] Update
---
.gitignore | 1 -
generate_dataset.py | 20 +-
geocoder_app.py | 1 +
lib/geocoder/our_geocoder.py | 4 +-
notebooks/BERT.ipynb | 9394 +++++++++++++++++
notebooks/ambiguity.ipynb | 387 +
notebooks/baseline.ipynb | 1773 ++++
notebooks/disambiguate.ipynb | 143 +
notebooks/generate_ambiguous_dataset.ipynb | 273 +
notebooks/heuristic.ipynb | 407 +
notebooks/tobert.ipynb | 297 +
scripts/embeddingngram.py | 59 -
scripts/extract_pages_of_interest.py | 48 +
scripts/generate_cooc_geocoding_dataset.py | 41 -
scripts/get_all_adjacency_rel.py | 88 -
scripts/gethealpix.py | 32 -
.../{randoludo.py => rando_ludo_geocoding.py} | 0
templates/pair_topo.html | 2 +-
18 files changed, 12739 insertions(+), 231 deletions(-)
create mode 100644 notebooks/BERT.ipynb
create mode 100644 notebooks/ambiguity.ipynb
create mode 100644 notebooks/baseline.ipynb
create mode 100644 notebooks/disambiguate.ipynb
create mode 100644 notebooks/generate_ambiguous_dataset.ipynb
create mode 100644 notebooks/heuristic.ipynb
create mode 100644 notebooks/tobert.ipynb
delete mode 100644 scripts/embeddingngram.py
create mode 100644 scripts/extract_pages_of_interest.py
delete mode 100644 scripts/generate_cooc_geocoding_dataset.py
delete mode 100644 scripts/get_all_adjacency_rel.py
delete mode 100644 scripts/gethealpix.py
rename scripts/{randoludo.py => rando_ludo_geocoding.py} (100%)
diff --git a/.gitignore b/.gitignore
index 2f0ed0e..92da221 100644
--- a/.gitignore
+++ b/.gitignore
@@ -133,7 +133,6 @@ dmypy.json
data/*
deprecated/*
*.ipynb_checkpoints
-notebooks/*
outputs*
temp/*
WikipediaExtract/*
diff --git a/generate_dataset.py b/generate_dataset.py
index 788fe37..bf1f8af 100644
--- a/generate_dataset.py
+++ b/generate_dataset.py
@@ -19,6 +19,7 @@ parser.add_argument("--adj-sampling", default=4, type=int)
parser.add_argument("--adj-nside", default=128, type=int)
parser.add_argument("--split-nside", default=128, type=int)
parser.add_argument("--split-method", default="per_pair", type=str, choices="per_pair per_place".split())
+parser.add_argument("--no-sampling", action="store_true")
args = parser.parse_args()#("../data/geonamesData/FR.txt ../data/wikipedia/cooccurrence_FR.txt ../data/geonamesData/hierarchy.txt".split())
@@ -40,8 +41,7 @@ wikipedia_data["geonameid"] = np.arange(min_id, max_id)
geonames_data["adj_split"] = geonames_data.apply(lambda x: latlon2healpix(x.latitude, x.longitude, args.adj_nside),
axis=1)
-
-def get_adjacent_pairs(dataframe, sampling_nb):
+def get_adjacent_pairs(dataframe, sampling_nb=4,no_sampling=False):
"""
Return pairs of place toponyms that are adjacent geographicaly.
Parameters
@@ -62,12 +62,15 @@ def get_adjacent_pairs(dataframe, sampling_nb):
topo_prin = row["name"]
lat, lon = row.latitude, row.longitude
within_cell = dataframe[dataframe.adj_split == healpix_cell]["name"].values
- selected = np.random.choice(within_cell, sampling_nb)
+ if not no_sampling:
+ selected = np.random.choice(within_cell, sampling_nb)
+ else:
+ selected = within_cell
new_pairs.extend([[row.geonameid, topo_prin, sel, lat, lon] for sel in selected])
return new_pairs
-def get_cooccurrence_pairs(dataframe, sampling_nb):
+def get_cooccurrence_pairs(dataframe, sampling_nb=4,no_sampling=False):
"""
Return pairs of place toponyms where toponyms appears in a same wikipedia page
Parameters
@@ -83,7 +86,10 @@ def get_cooccurrence_pairs(dataframe, sampling_nb):
[[ID,place toponym,adjacent place toponym, latitude, longitude],...]
"""
new_pairs = []
- dataframe["interlinks"] = dataframe.interlinks.apply(lambda x: np.random.choice(x.split("|"), sampling_nb))
+ if not no_sampling:
+ dataframe["interlinks"] = dataframe.interlinks.apply(lambda x: np.random.choice(x.split("|"), sampling_nb))
+ else:
+ dataframe["interlinks"] = dataframe.interlinks.apply(lambda x: x.split("|"))
for ix, row in tqdm(dataframe.iterrows(), total=len(dataframe), desc="Get Cooccurrent Toponym Pairs"):
topo_prin = row.title
lat, lon = row.latitude, row.longitude
@@ -115,9 +121,9 @@ def get_inclusion_pairs(geoname_df, hierarchy_df):
return [[p[0], id_label[p[0]], id_label[p[1]], id_lat[p[0]], id_lon[p[0]]] for p in pairs_id]
# EXTRACT PAIRS FROM INPUT DATA
-cooc_pairs = pd.DataFrame(get_cooccurrence_pairs(wikipedia_data, args.cooc_sampling),
+cooc_pairs = pd.DataFrame(get_cooccurrence_pairs(wikipedia_data, args.cooc_sampling,args.no_sampling),
columns="ID toponym toponym_context latitude longitude".split())
-adjacent_pairs = pd.DataFrame(get_adjacent_pairs(geonames_data, args.adj_sampling),
+adjacent_pairs = pd.DataFrame(get_adjacent_pairs(geonames_data, args.adj_sampling,args.no_sampling),
columns="ID toponym toponym_context latitude longitude".split())
inclusion_pairs = pd.DataFrame(get_inclusion_pairs(geonames_data, geonames_hierarchy_data),
columns="ID toponym toponym_context latitude longitude".split())
diff --git a/geocoder_app.py b/geocoder_app.py
index 81d4a1e..b3c54cf 100644
--- a/geocoder_app.py
+++ b/geocoder_app.py
@@ -11,6 +11,7 @@ dict_model = {
"FR_C":("./outputs/FR_MODEL_2/FR.txt_100_4_100__C.h5","./outputs/FR_MODEL_2/FR.txt_100_4_100__C_index"),
"FR_AC":("./outputs/FR_MODEL_2/FR.txt_100_4_100__A_C.h5","./outputs/FR_MODEL_2/FR.txt_100_4_100__A_C_index"),
"FR_IC":("./outputs/FR_MODEL_2/FR.txt_100_4_100__I_C.h5","./outputs/FR_MODEL_2/FR.txt_100_4_100__I_C_index"),
+ "FR_AIC_512":("outputs/FR_MODEL_2/nside_512/FR_nside_512/FR_nside_512_4_100_A_I_P.h5.part","outputs/FR_MODEL_2/nside_512/FR_nside_512/FR_nside_512_4_100_A_I_P_index"),
"GB_AIC":("./outputs/GB_MODEL_2/GB.txt_100_4_100__A_I_C.h5","./outputs/GB_MODEL_2/GB.txt_100_4_100__A_I_C_index"),
"GB_C":("./outputs/GB_MODEL_2/GB.txt_100_4_100__C.h5","./outputs/GB_MODEL_2/GB.txt_100_4_100__C_index"),
diff --git a/lib/geocoder/our_geocoder.py b/lib/geocoder/our_geocoder.py
index 0345ea4..9a9a9ae 100644
--- a/lib/geocoder/our_geocoder.py
+++ b/lib/geocoder/our_geocoder.py
@@ -15,7 +15,7 @@ from tensorflow.python.keras.models import load_model
# CUSTOM LIB
from lib.word_index import WordIndex
from lib.ngram_index import NgramIndex
-from lib.utils_geo import haversine_tf_1circle
+from lib.utils_geo import haversine_tf_1circle,accuracy_k
import stanza
@@ -33,7 +33,7 @@ class Geocoder(object):
if you want an interactive map using leafletJS, set to True the `interactive_map` parameter of `Geocoder.plot_coord()`
"""
def __init__(self,keras_model_fn,ngram_index_file,word_index=False):
- self.keras_model = load_model(keras_model_fn,custom_objects={"loss":haversine_tf_1circle},compile=False)#custom_objects={"accuracy_at_k_lat":lat_accuracy(),"accuracy_at_k_lon":lon_accuracy()})
+ self.keras_model = load_model(keras_model_fn,custom_objects={"haversine_tf_1circle":haversine_tf_1circle,"compute_metric":accuracy_k()},compile=False)#custom_objects={"accuracy_at_k_lat":lat_accuracy(),"accuracy_at_k_lon":lon_accuracy()})
if not word_index:
self.ngram_encoder = NgramIndex.load(ngram_index_file)
else:
diff --git a/notebooks/BERT.ipynb b/notebooks/BERT.ipynb
new file mode 100644
index 0000000..10c1f37
--- /dev/null
+++ b/notebooks/BERT.ipynb
@@ -0,0 +1,9394 @@
+{
+ "nbformat": 4,
+ "nbformat_minor": 0,
+ "metadata": {
+ "colab": {
+ "name": "Copie de BERT_cased_uncased_Fine_Tuning_Sentence_Classification_Geo_22_06.ipynb",
+ "provenance": [],
+ "collapsed_sections": [],
+ "toc_visible": true
+ },
+ "kernelspec": {
+ "name": "python3",
+ "display_name": "Python 3"
+ },
+ "accelerator": "GPU",
+ "widgets": {
+ "application/vnd.jupyter.widget-state+json": {
+ "f140816dfbff4cbaabada70d5229c1df": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HBoxModel",
+ "state": {
+ "_view_name": "HBoxView",
+ "_dom_classes": [],
+ "_model_name": "HBoxModel",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "box_style": "",
+ "layout": "IPY_MODEL_97ef177e14234ae58822284b1888c988",
+ "_model_module": "@jupyter-widgets/controls",
+ "children": [
+ "IPY_MODEL_23f60c963e724de7aed0444667462b0f",
+ "IPY_MODEL_72826ad45501458694a6e14e279d1883"
+ ]
+ }
+ },
+ "97ef177e14234ae58822284b1888c988": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "23f60c963e724de7aed0444667462b0f": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "FloatProgressModel",
+ "state": {
+ "_view_name": "ProgressView",
+ "style": "IPY_MODEL_6007d90f8fa2410e89475f0c97a72e17",
+ "_dom_classes": [],
+ "description": "Downloading: 100%",
+ "_model_name": "FloatProgressModel",
+ "bar_style": "success",
+ "max": 995526,
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": 995526,
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "orientation": "horizontal",
+ "min": 0,
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_3455ef60f99546ba9b55191423e2ce2f"
+ }
+ },
+ "72826ad45501458694a6e14e279d1883": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HTMLModel",
+ "state": {
+ "_view_name": "HTMLView",
+ "style": "IPY_MODEL_c2408942aa264664a51c96c4eb932b54",
+ "_dom_classes": [],
+ "description": "",
+ "_model_name": "HTMLModel",
+ "placeholder": "​",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": " 996k/996k [00:01<00:00, 807kB/s]",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_df0ac693530647389f25a6dfb84b9b50"
+ }
+ },
+ "6007d90f8fa2410e89475f0c97a72e17": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "ProgressStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "ProgressStyleModel",
+ "description_width": "initial",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "bar_color": null,
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "3455ef60f99546ba9b55191423e2ce2f": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "c2408942aa264664a51c96c4eb932b54": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "DescriptionStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "DescriptionStyleModel",
+ "description_width": "",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "df0ac693530647389f25a6dfb84b9b50": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "d3ac76358c3246f8897b55741031a3e5": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HBoxModel",
+ "state": {
+ "_view_name": "HBoxView",
+ "_dom_classes": [],
+ "_model_name": "HBoxModel",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "box_style": "",
+ "layout": "IPY_MODEL_38e2da0938504f90bd18c68d7e2ffc51",
+ "_model_module": "@jupyter-widgets/controls",
+ "children": [
+ "IPY_MODEL_2e298550f61b4d08af0a6497d8e33b5b",
+ "IPY_MODEL_62636381fe554245bb5efc22a509ffc1"
+ ]
+ }
+ },
+ "38e2da0938504f90bd18c68d7e2ffc51": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "2e298550f61b4d08af0a6497d8e33b5b": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "FloatProgressModel",
+ "state": {
+ "_view_name": "ProgressView",
+ "style": "IPY_MODEL_5ccfbc88b71d4f15ba6bf374bb1007ed",
+ "_dom_classes": [],
+ "description": "Downloading: 100%",
+ "_model_name": "FloatProgressModel",
+ "bar_style": "success",
+ "max": 625,
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": 625,
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "orientation": "horizontal",
+ "min": 0,
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_a053a05834ae45838dca81a8ed4b0a27"
+ }
+ },
+ "62636381fe554245bb5efc22a509ffc1": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HTMLModel",
+ "state": {
+ "_view_name": "HTMLView",
+ "style": "IPY_MODEL_474a715d1a054ec29547c1b54560f7f6",
+ "_dom_classes": [],
+ "description": "",
+ "_model_name": "HTMLModel",
+ "placeholder": "​",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": " 625/625 [00:00<00:00, 3.46kB/s]",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_ed377908b70442429d666e123f526e0e"
+ }
+ },
+ "5ccfbc88b71d4f15ba6bf374bb1007ed": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "ProgressStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "ProgressStyleModel",
+ "description_width": "initial",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "bar_color": null,
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "a053a05834ae45838dca81a8ed4b0a27": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "474a715d1a054ec29547c1b54560f7f6": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "DescriptionStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "DescriptionStyleModel",
+ "description_width": "",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "ed377908b70442429d666e123f526e0e": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "86290b6fa15844388475ac4a10d6a3aa": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HBoxModel",
+ "state": {
+ "_view_name": "HBoxView",
+ "_dom_classes": [],
+ "_model_name": "HBoxModel",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "box_style": "",
+ "layout": "IPY_MODEL_e77e34463f3f4de5abd1e8e9a28c4da7",
+ "_model_module": "@jupyter-widgets/controls",
+ "children": [
+ "IPY_MODEL_0a1e169b0359416e93e9f2e5c6c7b1bb",
+ "IPY_MODEL_7ad677239085495892d73c33ce07af73"
+ ]
+ }
+ },
+ "e77e34463f3f4de5abd1e8e9a28c4da7": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "0a1e169b0359416e93e9f2e5c6c7b1bb": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "FloatProgressModel",
+ "state": {
+ "_view_name": "ProgressView",
+ "style": "IPY_MODEL_fbf9d95bd65545cc8c51f3a5d80aef59",
+ "_dom_classes": [],
+ "description": "Downloading: 100%",
+ "_model_name": "FloatProgressModel",
+ "bar_style": "success",
+ "max": 714314041,
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": 714314041,
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "orientation": "horizontal",
+ "min": 0,
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_f31e5ef9468e4fefb41ed01d2ecb337c"
+ }
+ },
+ "7ad677239085495892d73c33ce07af73": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HTMLModel",
+ "state": {
+ "_view_name": "HTMLView",
+ "style": "IPY_MODEL_084f12e8365d448fab0d70b7589182b7",
+ "_dom_classes": [],
+ "description": "",
+ "_model_name": "HTMLModel",
+ "placeholder": "​",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": " 714M/714M [10:19<00:00, 1.15MB/s]",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_c57df1509cd240f999d37e4a1bd5c075"
+ }
+ },
+ "fbf9d95bd65545cc8c51f3a5d80aef59": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "ProgressStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "ProgressStyleModel",
+ "description_width": "initial",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "bar_color": null,
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "f31e5ef9468e4fefb41ed01d2ecb337c": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "084f12e8365d448fab0d70b7589182b7": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "DescriptionStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "DescriptionStyleModel",
+ "description_width": "",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "c57df1509cd240f999d37e4a1bd5c075": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "ae4a115927af4738b9c3f2a8cb97a32d": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HBoxModel",
+ "state": {
+ "_view_name": "HBoxView",
+ "_dom_classes": [],
+ "_model_name": "HBoxModel",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "box_style": "",
+ "layout": "IPY_MODEL_da83613435044dfdb693f7cd6293898e",
+ "_model_module": "@jupyter-widgets/controls",
+ "children": [
+ "IPY_MODEL_6a0aa93d9b3841d7934cd07768b4e8d9",
+ "IPY_MODEL_3fd2e1b55945449e81a96e89b5e0015d"
+ ]
+ }
+ },
+ "da83613435044dfdb693f7cd6293898e": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "6a0aa93d9b3841d7934cd07768b4e8d9": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "FloatProgressModel",
+ "state": {
+ "_view_name": "ProgressView",
+ "style": "IPY_MODEL_7f7f5353878d4046b5a1e0b831f923b9",
+ "_dom_classes": [],
+ "description": "Downloading: 100%",
+ "_model_name": "FloatProgressModel",
+ "bar_style": "success",
+ "max": 871891,
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": 871891,
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "orientation": "horizontal",
+ "min": 0,
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_49650cd4a0664497839d492c0911acc7"
+ }
+ },
+ "3fd2e1b55945449e81a96e89b5e0015d": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HTMLModel",
+ "state": {
+ "_view_name": "HTMLView",
+ "style": "IPY_MODEL_28aef9ee16b24d4d94ff728d7e7f7b6a",
+ "_dom_classes": [],
+ "description": "",
+ "_model_name": "HTMLModel",
+ "placeholder": "​",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": " 872k/872k [00:01<00:00, 715kB/s]",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_b8f06024ff9f4d24aa087732be6baeab"
+ }
+ },
+ "7f7f5353878d4046b5a1e0b831f923b9": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "ProgressStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "ProgressStyleModel",
+ "description_width": "initial",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "bar_color": null,
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "49650cd4a0664497839d492c0911acc7": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "28aef9ee16b24d4d94ff728d7e7f7b6a": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "DescriptionStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "DescriptionStyleModel",
+ "description_width": "",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "b8f06024ff9f4d24aa087732be6baeab": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "8493b7b6fefb4462bb064057c2bbdefc": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HBoxModel",
+ "state": {
+ "_view_name": "HBoxView",
+ "_dom_classes": [],
+ "_model_name": "HBoxModel",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "box_style": "",
+ "layout": "IPY_MODEL_9a1ff0584fa045dc82c15ca21977581a",
+ "_model_module": "@jupyter-widgets/controls",
+ "children": [
+ "IPY_MODEL_1030b0c14bdd4bdbb79d7cd3a04a192a",
+ "IPY_MODEL_167283864c1543c7be6adb2bd0434a02"
+ ]
+ }
+ },
+ "9a1ff0584fa045dc82c15ca21977581a": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "1030b0c14bdd4bdbb79d7cd3a04a192a": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "FloatProgressModel",
+ "state": {
+ "_view_name": "ProgressView",
+ "style": "IPY_MODEL_3a77f366f755438f933481739e8d4e2d",
+ "_dom_classes": [],
+ "description": "Downloading: 100%",
+ "_model_name": "FloatProgressModel",
+ "bar_style": "success",
+ "max": 625,
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": 625,
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "orientation": "horizontal",
+ "min": 0,
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_bb62a9b6ee864446a1f3c16b07e04420"
+ }
+ },
+ "167283864c1543c7be6adb2bd0434a02": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HTMLModel",
+ "state": {
+ "_view_name": "HTMLView",
+ "style": "IPY_MODEL_b60dbc236d89455ca868903951919194",
+ "_dom_classes": [],
+ "description": "",
+ "_model_name": "HTMLModel",
+ "placeholder": "​",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": " 625/625 [00:00<00:00, 3.93kB/s]",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_66c416fdc3774f948ff0e1ea814eaa67"
+ }
+ },
+ "3a77f366f755438f933481739e8d4e2d": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "ProgressStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "ProgressStyleModel",
+ "description_width": "initial",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "bar_color": null,
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "bb62a9b6ee864446a1f3c16b07e04420": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "b60dbc236d89455ca868903951919194": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "DescriptionStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "DescriptionStyleModel",
+ "description_width": "",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "66c416fdc3774f948ff0e1ea814eaa67": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "a89d900918634fbd84db67e6c0eacbd3": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HBoxModel",
+ "state": {
+ "_view_name": "HBoxView",
+ "_dom_classes": [],
+ "_model_name": "HBoxModel",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "box_style": "",
+ "layout": "IPY_MODEL_fd849fa776164444b3d1d1c9f0a5ecff",
+ "_model_module": "@jupyter-widgets/controls",
+ "children": [
+ "IPY_MODEL_487f4e58a4634112b021276556118d49",
+ "IPY_MODEL_2156a8fec63848b285dff96898bc84d8"
+ ]
+ }
+ },
+ "fd849fa776164444b3d1d1c9f0a5ecff": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "487f4e58a4634112b021276556118d49": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "FloatProgressModel",
+ "state": {
+ "_view_name": "ProgressView",
+ "style": "IPY_MODEL_d906ff1cb67040c0b91d7928e1b123c4",
+ "_dom_classes": [],
+ "description": "Downloading: 100%",
+ "_model_name": "FloatProgressModel",
+ "bar_style": "success",
+ "max": 672271273,
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": 672271273,
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "orientation": "horizontal",
+ "min": 0,
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_81bf0d3185c340218382cab00eb91fe5"
+ }
+ },
+ "2156a8fec63848b285dff96898bc84d8": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "HTMLModel",
+ "state": {
+ "_view_name": "HTMLView",
+ "style": "IPY_MODEL_56626b5f15df4a9492424e20b7f31012",
+ "_dom_classes": [],
+ "description": "",
+ "_model_name": "HTMLModel",
+ "placeholder": "​",
+ "_view_module": "@jupyter-widgets/controls",
+ "_model_module_version": "1.5.0",
+ "value": " 672M/672M [00:40<00:00, 16.5MB/s]",
+ "_view_count": null,
+ "_view_module_version": "1.5.0",
+ "description_tooltip": null,
+ "_model_module": "@jupyter-widgets/controls",
+ "layout": "IPY_MODEL_939183280d4a48cb8825073ece64268d"
+ }
+ },
+ "d906ff1cb67040c0b91d7928e1b123c4": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "ProgressStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "ProgressStyleModel",
+ "description_width": "initial",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "bar_color": null,
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "81bf0d3185c340218382cab00eb91fe5": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ },
+ "56626b5f15df4a9492424e20b7f31012": {
+ "model_module": "@jupyter-widgets/controls",
+ "model_name": "DescriptionStyleModel",
+ "state": {
+ "_view_name": "StyleView",
+ "_model_name": "DescriptionStyleModel",
+ "description_width": "",
+ "_view_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.5.0",
+ "_view_count": null,
+ "_view_module_version": "1.2.0",
+ "_model_module": "@jupyter-widgets/controls"
+ }
+ },
+ "939183280d4a48cb8825073ece64268d": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "state": {
+ "_view_name": "LayoutView",
+ "grid_template_rows": null,
+ "right": null,
+ "justify_content": null,
+ "_view_module": "@jupyter-widgets/base",
+ "overflow": null,
+ "_model_module_version": "1.2.0",
+ "_view_count": null,
+ "flex_flow": null,
+ "width": null,
+ "min_width": null,
+ "border": null,
+ "align_items": null,
+ "bottom": null,
+ "_model_module": "@jupyter-widgets/base",
+ "top": null,
+ "grid_column": null,
+ "overflow_y": null,
+ "overflow_x": null,
+ "grid_auto_flow": null,
+ "grid_area": null,
+ "grid_template_columns": null,
+ "flex": null,
+ "_model_name": "LayoutModel",
+ "justify_items": null,
+ "grid_row": null,
+ "max_height": null,
+ "align_content": null,
+ "visibility": null,
+ "align_self": null,
+ "height": null,
+ "min_height": null,
+ "padding": null,
+ "grid_auto_rows": null,
+ "grid_gap": null,
+ "max_width": null,
+ "order": null,
+ "_view_module_version": "1.2.0",
+ "grid_template_areas": null,
+ "object_position": null,
+ "object_fit": null,
+ "grid_auto_columns": null,
+ "margin": null,
+ "display": null,
+ "left": null
+ }
+ }
+ }
+ }
+ },
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "RX_ZDhicpHkV"
+ },
+ "source": [
+ "# 1. Setup"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "nSU7yERLP_66"
+ },
+ "source": [
+ "## 1.1. Using Colab GPU for Training\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "GI0iOY8zvZzL"
+ },
+ "source": [
+ "\n",
+ "Google Colab offers free GPUs and TPUs! Since I'll be training a large neural network it's best to take advantage of this (in this case I'll attach a GPU), otherwise training will take a very long time.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "DEfSbAA4QHas",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "739df07b-bc9d-4c2a-8a2e-b272483af063"
+ },
+ "source": [
+ "import tensorflow as tf\n",
+ "\n",
+ "# Get the GPU device name.\n",
+ "device_name = tf.test.gpu_device_name()\n",
+ "\n",
+ "# The device name should look like the following:\n",
+ "if device_name == '/device:GPU:0':\n",
+ " print('Found GPU at: {}'.format(device_name))\n",
+ "else:\n",
+ " raise SystemError('GPU device not found')"
+ ],
+ "execution_count": 1,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Found GPU at: /device:GPU:0\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "cqG7FzRVFEIv"
+ },
+ "source": [
+ "In order for torch to use the GPU, I need to identify and specify the GPU as the device. Later, in my training loop, I will load data onto the device. "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "oYsV4H8fCpZ-",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "54dc1665-1dea-4114-fd2a-526b279a3b33"
+ },
+ "source": [
+ "# We need the GPU: Tesla P100-PCIE-16GB\n",
+ "import torch\n",
+ "\n",
+ "# If there's a GPU available...\n",
+ "if torch.cuda.is_available(): \n",
+ "\n",
+ " # Tell PyTorch to use the GPU. \n",
+ " device = torch.device(\"cuda\")\n",
+ "\n",
+ " print('There are %d GPU(s) available.' % torch.cuda.device_count())\n",
+ "\n",
+ " print('We will use the GPU:', torch.cuda.get_device_name(0))\n",
+ "\n",
+ "# If not...\n",
+ "else:\n",
+ " print('No GPU available, using the CPU instead.')\n",
+ " device = torch.device(\"cpu\")"
+ ],
+ "execution_count": 2,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "There are 1 GPU(s) available.\n",
+ "We will use the GPU: Tesla T4\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "0NmMdkZO8R6q",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "74fd46ac-9990-4d3f-f415-bc0352597c63"
+ },
+ "source": [
+ "!pip install transformers"
+ ],
+ "execution_count": 3,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Collecting transformers\n",
+ "\u001b[?25l Downloading https://files.pythonhosted.org/packages/3a/83/e74092e7f24a08d751aa59b37a9fc572b2e4af3918cb66f7766c3affb1b4/transformers-3.5.1-py3-none-any.whl (1.3MB)\n",
+ "\u001b[K |████████████████████████████████| 1.3MB 8.6MB/s \n",
+ "\u001b[?25hCollecting sentencepiece==0.1.91\n",
+ "\u001b[?25l Downloading https://files.pythonhosted.org/packages/d4/a4/d0a884c4300004a78cca907a6ff9a5e9fe4f090f5d95ab341c53d28cbc58/sentencepiece-0.1.91-cp36-cp36m-manylinux1_x86_64.whl (1.1MB)\n",
+ "\u001b[K |████████████████████████████████| 1.1MB 29.6MB/s \n",
+ "\u001b[?25hRequirement already satisfied: packaging in /usr/local/lib/python3.6/dist-packages (from transformers) (20.4)\n",
+ "Collecting sacremoses\n",
+ "\u001b[?25l Downloading https://files.pythonhosted.org/packages/7d/34/09d19aff26edcc8eb2a01bed8e98f13a1537005d31e95233fd48216eed10/sacremoses-0.0.43.tar.gz (883kB)\n",
+ "\u001b[K |████████████████████████████████| 890kB 19.4MB/s \n",
+ "\u001b[?25hRequirement already satisfied: dataclasses; python_version < \"3.7\" in /usr/local/lib/python3.6/dist-packages (from transformers) (0.7)\n",
+ "Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.6/dist-packages (from transformers) (2019.12.20)\n",
+ "Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.6/dist-packages (from transformers) (4.41.1)\n",
+ "Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from transformers) (2.23.0)\n",
+ "Requirement already satisfied: filelock in /usr/local/lib/python3.6/dist-packages (from transformers) (3.0.12)\n",
+ "Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from transformers) (1.18.5)\n",
+ "Requirement already satisfied: protobuf in /usr/local/lib/python3.6/dist-packages (from transformers) (3.12.4)\n",
+ "Collecting tokenizers==0.9.3\n",
+ "\u001b[?25l Downloading https://files.pythonhosted.org/packages/4c/34/b39eb9994bc3c999270b69c9eea40ecc6f0e97991dba28282b9fd32d44ee/tokenizers-0.9.3-cp36-cp36m-manylinux1_x86_64.whl (2.9MB)\n",
+ "\u001b[K |████████████████████████████████| 2.9MB 52.3MB/s \n",
+ "\u001b[?25hRequirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from packaging->transformers) (1.15.0)\n",
+ "Requirement already satisfied: pyparsing>=2.0.2 in /usr/local/lib/python3.6/dist-packages (from packaging->transformers) (2.4.7)\n",
+ "Requirement already satisfied: click in /usr/local/lib/python3.6/dist-packages (from sacremoses->transformers) (7.1.2)\n",
+ "Requirement already satisfied: joblib in /usr/local/lib/python3.6/dist-packages (from sacremoses->transformers) (0.17.0)\n",
+ "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (3.0.4)\n",
+ "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (2.10)\n",
+ "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (1.24.3)\n",
+ "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (2020.6.20)\n",
+ "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf->transformers) (50.3.2)\n",
+ "Building wheels for collected packages: sacremoses\n",
+ " Building wheel for sacremoses (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
+ " Created wheel for sacremoses: filename=sacremoses-0.0.43-cp36-none-any.whl size=893257 sha256=ced66aa7ed83946e7dc652ebd7f087d89d04ed88f3550f3c31ba3a4f85fff43c\n",
+ " Stored in directory: /root/.cache/pip/wheels/29/3c/fd/7ce5c3f0666dab31a50123635e6fb5e19ceb42ce38d4e58f45\n",
+ "Successfully built sacremoses\n",
+ "Installing collected packages: sentencepiece, sacremoses, tokenizers, transformers\n",
+ "Successfully installed sacremoses-0.0.43 sentencepiece-0.1.91 tokenizers-0.9.3 transformers-3.5.1\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "lxddqmruamSj"
+ },
+ "source": [
+ "The code in this notebook is actually a simplified version of the [run_glue.py](https://github.com/huggingface/transformers/blob/master/examples/run_glue.py) example script from huggingface.\n",
+ "\n",
+ "`run_glue.py` is a helpful utility which allows you to pick which GLUE benchmark task you want to run on, and which pre-trained model you want to use (you can see the list of possible models [here](https://github.com/huggingface/transformers/blob/e6cff60b4cbc1158fbd6e4a1c3afda8dc224f566/examples/run_glue.py#L69)). It also supports using either the CPU, a single GPU, or multiple GPUs. It even supports using 16-bit precision if you want further speed up.\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "guw6ZNtaswKc"
+ },
+ "source": [
+ "# 2. Loading Dataset\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "4JrUHXms16cn"
+ },
+ "source": [
+ "## 2.1. Download & Extract"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "z-4rTOcWwlbX"
+ },
+ "source": [
+ "import pandas as pd"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "dPcKbwzMw4KQ",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "f4fb8ce3-19a5-4d01-8459-174cafe44678"
+ },
+ "source": [
+ "Geo_analyse = pd.read_csv(\"/content/train_moins_sample.csv\")\n",
+ "\n",
+ "Geo_analyse.isnull().values.any()\n",
+ "\n",
+ "Geo_analyse.shape"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "(2317, 2)"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 5
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "oQUy9Tat2EF_"
+ },
+ "source": [
+ "## 2.2. Parse"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "H_LpQfzCn9_o"
+ },
+ "source": [
+ "Here are five sentences of the dataset"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "blqIvQaQncdJ",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 206
+ },
+ "outputId": "d8773029-d9bc-4789-b1a8-cb4763b3d3e3"
+ },
+ "source": [
+ "Geo_analyse.head(5)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>sentences</th>\n",
+ " <th>labels</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>— Comme tu voudras, répondit-elle.</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>Ce n’est plus moi, c’est elle qui couche avec ...</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>Un parent, je crois.</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>je m’y attendais un peu.</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>Comment la faire taire?</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " sentences labels\n",
+ "0 — Comme tu voudras, répondit-elle. 0\n",
+ "1 Ce n’est plus moi, c’est elle qui couche avec ... 0\n",
+ "2 Un parent, je crois. 0\n",
+ "3 je m’y attendais un peu. 0\n",
+ "4 Comment la faire taire? 0"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 8
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "4SMZ5T5Imhlx"
+ },
+ "source": [
+ "\n",
+ "\n",
+ "Let's extract the sentences and labels of our training set as numpy ndarrays."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "GuE5BqICAne2"
+ },
+ "source": [
+ "# Get the lists of sentences and their labels.\n",
+ "sentences = Geo_analyse.sentences.values\n",
+ "labels = Geo_analyse.labels.values"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "ex5O1eV-Pfct"
+ },
+ "source": [
+ "# 3. Tokenization & Input Formatting\n",
+ "\n",
+ "In this section, we'll transform our dataset into the format that BERT can be trained on."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "-8kEDRvShcU5"
+ },
+ "source": [
+ "## 3.1. BERT Tokenizer"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "bWOPOyWghJp2"
+ },
+ "source": [
+ "\n",
+ "To feed our text to BERT, it must be split into tokens, and then these tokens must be mapped to their index in the tokenizer vocabulary.\n",
+ "\n",
+ "The tokenization must be performed by the tokenizer included with BERT--the below cell will download this for us. We'll be using the \"multilingual cased\" version here.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Z474sSC6oe7A",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 86,
+ "referenced_widgets": [
+ "f140816dfbff4cbaabada70d5229c1df",
+ "97ef177e14234ae58822284b1888c988",
+ "23f60c963e724de7aed0444667462b0f",
+ "72826ad45501458694a6e14e279d1883",
+ "6007d90f8fa2410e89475f0c97a72e17",
+ "3455ef60f99546ba9b55191423e2ce2f",
+ "c2408942aa264664a51c96c4eb932b54",
+ "df0ac693530647389f25a6dfb84b9b50"
+ ]
+ },
+ "outputId": "0a79b322-539b-4e81-aacd-6b429c5e2a1c"
+ },
+ "source": [
+ "from transformers import BertTokenizer\n",
+ "\n",
+ "# Load the BERT tokenizer.\n",
+ "print('Loading BERT tokenizer...')\n",
+ "tokenizer = BertTokenizer.from_pretrained('bert-base-multilingual-cased',do_lower_case=False)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Loading BERT tokenizer...\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "f140816dfbff4cbaabada70d5229c1df",
+ "version_minor": 0,
+ "version_major": 2
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=995526.0, style=ProgressStyle(descripti…"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "dFzmtleW6KmJ"
+ },
+ "source": [
+ "Apply the tokenizer to one sentence just to see the output.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "dLIbudgfh6F0",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 73
+ },
+ "outputId": "b5532647-7220-4f5f-eec5-c3af8478c6e0"
+ },
+ "source": [
+ "# Print the original sentence.\n",
+ "print(' Original: ', sentences[0])\n",
+ "\n",
+ "# Print the sentence split into tokens.\n",
+ "print('Tokenized: ', tokenizer.tokenize(sentences[0]))\n",
+ "\n",
+ "# Print the sentence mapped to token ids.\n",
+ "print('Token IDs: ', tokenizer.convert_tokens_to_ids(tokenizer.tokenize(sentences[0])))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ " Original: — Comme tu voudras, répondit-elle.\n",
+ "Tokenized: ['[UNK]', 'Comme', 'tu', 'vo', '##udra', '##s', ',', 'répond', '##it', '-', 'elle', '.']\n",
+ "Token IDs: [100, 27113, 13055, 12556, 81216, 10107, 117, 108811, 10486, 118, 11117, 119]\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "WeNIc4auFUdF"
+ },
+ "source": [
+ "When I actually convert all of my sentences, I'll use the `tokenize.encode` function to handle both steps, rather than calling `tokenize` and `convert_tokens_to_ids` separately. \n",
+ "\n",
+ "Before I can do that, though, I need to talk about some of BERT's formatting requirements."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "viKGCCh8izww"
+ },
+ "source": [
+ "## 3.2. Required Formatting"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "yDcqNlvVhL5W"
+ },
+ "source": [
+ "The above code left out a few required formatting steps that I'll look at here.\n",
+ "\n",
+ "*Side Note: The input format to BERT seems \"over-specified\" to me... We are required to give it a number of pieces of information which seem redundant, or like they could easily be inferred from the data without us explicity providing it. But it is what it is, and I suspect it will make more sense once I have a deeper understanding of the BERT internals.*\n",
+ "\n",
+ "I am required to:\n",
+ "1. Add special tokens to the start and end of each sentence.\n",
+ "2. Pad & truncate all sentences to a single constant length.\n",
+ "3. Explicitly differentiate real tokens from padding tokens with the \"attention mask\".\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "V6mceWWOjZnw"
+ },
+ "source": [
+ "### Special Tokens\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Ykk0P9JiKtVe"
+ },
+ "source": [
+ "\n",
+ "**`[SEP]`**\n",
+ "\n",
+ "At the end of every sentence, I need to append the special `[SEP]` token. \n",
+ "\n",
+ "This token is an artifact of two-sentence tasks, where BERT is given two separate sentences and asked to determine something (e.g., can the answer to the question in sentence A be found in sentence B?). "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "86C9objaKu8f"
+ },
+ "source": [
+ "**`[CLS]`**\n",
+ "\n",
+ "For classification tasks, I must prepend the special `[CLS]` token to the beginning of every sentence.\n",
+ "\n",
+ "This token has special significance. BERT consists of 12 Transformer layers. Each transformer takes in a list of token embeddings, and produces the same number of embeddings on the output (but with the feature values changed, of course!).\n",
+ "\n",
+ "\n",
+ "\n",
+ "On the output of the final (12th) transformer, *only the first embedding (corresponding to the [CLS] token) is used by the classifier*.\n",
+ "\n",
+ "> \"The first token of every sequence is always a special classification token (`[CLS]`). The final hidden state\n",
+ "corresponding to this token is used as the aggregate sequence representation for classification\n",
+ "tasks.\" (from the [BERT paper](https://arxiv.org/pdf/1810.04805.pdf))\n",
+ "\n",
+ "Also, because BERT is trained to only use this [CLS] token for classification, we know that the model has been motivated to encode everything it needs for the classification step into that single 768-value embedding vector.\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "u51v0kFxeteu"
+ },
+ "source": [
+ "### Sentence Length & Attention Mask\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "qPNuwqZVK3T6"
+ },
+ "source": [
+ "The sentences in my dataset obviously have varying lengths, so how does BERT handle this?\n",
+ "\n",
+ "BERT has two constraints:\n",
+ "1. All sentences must be padded or truncated to a single, fixed length.\n",
+ "2. The maximum sentence length is 512 tokens.\n",
+ "\n",
+ "Padding is done with a special `[PAD]` token, which is at index 0 in the BERT vocabulary. The below illustration demonstrates padding out to a \"MAX_LEN\" of 8 tokens.\n",
+ "\n",
+ "<img src=\"http://www.mccormickml.com/assets/BERT/padding_and_mask.png\" width=\"600\">\n",
+ "\n",
+ "The \"Attention Mask\" is simply an array of 1s and 0s indicating which tokens are padding and which aren't (seems kind of redundant, doesn't it?! Again, I don't currently know why).\n",
+ "\n",
+ "\n",
+ "\n",
+ "\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "l6w8elb-58GJ"
+ },
+ "source": [
+ "## 3.2. Sentences to IDs"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "1M296yz577fV"
+ },
+ "source": [
+ "The `tokenizer.encode` function combines multiple steps for us:\n",
+ "1. Split the sentence into tokens.\n",
+ "2. Add the special `[CLS]` and `[SEP]` tokens.\n",
+ "3. Map the tokens to their IDs."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "2bBdb3pt8LuQ",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "a334fcf0-1a03-438c-b060-94e4d1df4c10"
+ },
+ "source": [
+ "# Tokenize all of the sentences and map the tokens to thier word IDs.\n",
+ "input_ids = []\n",
+ "\n",
+ "# For every sentence...\n",
+ "for sent in sentences:\n",
+ " # `encode` will:\n",
+ " # (1) Tokenize the sentence.\n",
+ " # (2) Prepend the `[CLS]` token to the start.\n",
+ " # (3) Append the `[SEP]` token to the end.\n",
+ " # (4) Map tokens to their IDs.\n",
+ " encoded_sent = tokenizer.encode(\n",
+ " sent, # Sentence to encode.\n",
+ " add_special_tokens = True, # Add '[CLS]' and '[SEP]'\n",
+ "\n",
+ " # This function also supports truncation and conversion\n",
+ " # to pytorch tensors, but I need to do padding, so I\n",
+ " # can't use these features.\n",
+ " #max_length = 128, # Truncate all sentences.\n",
+ " #return_tensors = 'pt', # Return pytorch tensors.\n",
+ " )\n",
+ " \n",
+ " # Add the encoded sentence to the list.\n",
+ " input_ids.append(encoded_sent)\n",
+ "\n",
+ "# Print sentence 0, now as a list of IDs.\n",
+ "print('Original: ', sentences[0])\n",
+ "print('Token IDs:', input_ids[0])"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Original: — Comme tu voudras, répondit-elle.\n",
+ "Token IDs: [101, 100, 27113, 13055, 12556, 81216, 10107, 117, 108811, 10486, 118, 11117, 119, 102]\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "WhwCKszh6ych"
+ },
+ "source": [
+ "## 3.3. Padding & Truncating"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "xytsw1oIfnX0"
+ },
+ "source": [
+ "Pad and truncate my sequences so that they all have the same length, `MAX_LEN`."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "zqiWTDrn_nGB"
+ },
+ "source": [
+ "First, what's the maximum sentence length in our dataset?"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "JhUZO9vc_l6T",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "5b9ffcab-532a-42ec-82d5-89ea32daeda0"
+ },
+ "source": [
+ "print('Max sentence length: ', max([len(sen) for sen in input_ids]))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Max sentence length: 238\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "hp-54FcQ_p3h"
+ },
+ "source": [
+ "Given that, I choose MAX_LEN = 256 and apply the padding."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Cp9BPRd1tMIo",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 148
+ },
+ "outputId": "6e212358-38f9-4a29-b2de-18e28fc2e4fb"
+ },
+ "source": [
+ "# I'll borrow the `pad_sequences` utility function to do this.\n",
+ "from keras.preprocessing.sequence import pad_sequences\n",
+ "\n",
+ "# Set the maximum sequence length.\n",
+ "# I've chosen 256 somewhat arbitrarily. It's slightly larger than the\n",
+ "# maximum training sentence length of 238...\n",
+ "MAX_LEN = 256\n",
+ "\n",
+ "print('\\nPadding/truncating all sentences to %d values...' % MAX_LEN)\n",
+ "\n",
+ "print('\\nPadding token: \"{:}\", ID: {:}'.format(tokenizer.pad_token, tokenizer.pad_token_id))\n",
+ "\n",
+ "# Pad our input tokens with value 0.\n",
+ "# \"post\" indicates that we want to pad and truncate at the end of the sequence,\n",
+ "# as opposed to the beginning.\n",
+ "input_ids = pad_sequences(input_ids, maxlen=MAX_LEN, dtype=\"long\", \n",
+ " value=0, truncating=\"post\", padding=\"post\")\n",
+ "\n",
+ "print('\\nDone.')"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "\n",
+ "Padding/truncating all sentences to 256 values...\n",
+ "\n",
+ "Padding token: \"[PAD]\", ID: 0\n",
+ "\n",
+ "Done.\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "Using TensorFlow backend.\n"
+ ],
+ "name": "stderr"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "kDs-MYtYH8sL"
+ },
+ "source": [
+ "## 3.4. Attention Masks"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "KhGulL1pExCT"
+ },
+ "source": [
+ "The attention mask simply makes it explicit which tokens are actual words versus which are padding. \n",
+ "\n",
+ "The BERT vocabulary does not use the ID 0, so if a token ID is 0, then it's padding, and otherwise it's a real token."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "cDoC24LeEv3N"
+ },
+ "source": [
+ "# Create attention masks\n",
+ "attention_masks = []\n",
+ "\n",
+ "# For each sentence...\n",
+ "for sent in input_ids:\n",
+ " \n",
+ " # Create the attention mask.\n",
+ " # - If a token ID is 0, then it's padding, set the mask to 0.\n",
+ " # - If a token ID is > 0, then it's a real token, set the mask to 1.\n",
+ " att_mask = [int(token_id > 0) for token_id in sent]\n",
+ " \n",
+ " # Store the attention mask for this sentence.\n",
+ " attention_masks.append(att_mask)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "aRp4O7D295d_"
+ },
+ "source": [
+ "## 3.5. Training & Validation Split\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "qu0ao7p8rb06"
+ },
+ "source": [
+ "Divide the training set to use 90% for training and 10% for validation."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "aFbE-UHvsb7-"
+ },
+ "source": [
+ "# Use train_test_split to split data into train and validation sets for\n",
+ "# training\n",
+ "from sklearn.model_selection import train_test_split\n",
+ "\n",
+ "# Use 90% for training and 10% for validation.\n",
+ "train_inputs, validation_inputs, train_labels, validation_labels = train_test_split(input_ids, labels, \n",
+ " random_state=2018, test_size=0.1)\n",
+ "# Do the same for the masks.\n",
+ "train_masks, validation_masks, _, _ = train_test_split(attention_masks, labels,\n",
+ " random_state=2018, test_size=0.1)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "7LzSbTqW9_BR"
+ },
+ "source": [
+ "## 3.6. Converting to PyTorch Data Types"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "6p1uXczp-Je4"
+ },
+ "source": [
+ "My model expects PyTorch tensors rather than numpy.ndarrays, so convert all of our dataset variables."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "jw5K2A5Ko1RF"
+ },
+ "source": [
+ "# Convert all inputs and labels into torch tensors, the required datatype \n",
+ "# for my model.\n",
+ "train_inputs = torch.tensor(train_inputs)\n",
+ "validation_inputs = torch.tensor(validation_inputs)\n",
+ "\n",
+ "train_labels = torch.tensor(train_labels)\n",
+ "validation_labels = torch.tensor(validation_labels)\n",
+ "\n",
+ "train_masks = torch.tensor(train_masks)\n",
+ "validation_masks = torch.tensor(validation_masks)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "dD9i6Z2pG-sN"
+ },
+ "source": [
+ "I'll also create an iterator for the dataset using the torch DataLoader class. This helps save on memory during training because, unlike a for loop, with an iterator the entire dataset does not need to be loaded into memory."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "GEgLpFVlo1Z-"
+ },
+ "source": [
+ "from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler\n",
+ "\n",
+ "# The DataLoader needs to know the batch size for training, so I specify it here.\n",
+ "# For fine-tuning BERT on a specific task, the authors recommend a batch size of\n",
+ "# 16 or 32.\n",
+ "\n",
+ "batch_size = 32\n",
+ "\n",
+ "# Create the DataLoader for training set.\n",
+ "train_data = TensorDataset(train_inputs, train_masks, train_labels)\n",
+ "train_sampler = RandomSampler(train_data)\n",
+ "train_dataloader = DataLoader(train_data, sampler=train_sampler, batch_size=batch_size)\n",
+ "\n",
+ "# Create the DataLoader for validation set.\n",
+ "validation_data = TensorDataset(validation_inputs, validation_masks, validation_labels)\n",
+ "validation_sampler = SequentialSampler(validation_data)\n",
+ "validation_dataloader = DataLoader(validation_data, sampler=validation_sampler, batch_size=batch_size)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "8bwa6Rts-02-"
+ },
+ "source": [
+ "# 4. Train Classification Model"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "3xYQ3iLO08SX"
+ },
+ "source": [
+ "Now that input data is properly formatted, it's time to fine tune the BERT model. "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "D6TKgyUzPIQc"
+ },
+ "source": [
+ "## 4.1. BertForSequenceClassification"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "1sjzRT1V0zwm"
+ },
+ "source": [
+ "For this task, I first want to modify the pre-trained BERT model to give outputs for classification, and then I want to continue training the model on the dataset until that the entire model, end-to-end, is well-suited for my task. \n",
+ "\n",
+ "Thankfully, the huggingface pytorch implementation includes a set of interfaces designed for a variety of NLP tasks. Though these interfaces are all built on top of a trained BERT model, each has different top layers and output types designed to accomodate their specific NLP task. \n",
+ "\n",
+ "Here is the current list of classes provided for fine-tuning:\n",
+ "* BertModel\n",
+ "* BertForPreTraining\n",
+ "* BertForMaskedLM\n",
+ "* BertForNextSentencePrediction\n",
+ "* **BertForSequenceClassification** - The one I'll use.\n",
+ "* BertForTokenClassification\n",
+ "* BertForQuestionAnswering\n",
+ "\n",
+ "The documentation for these can be found under [here](https://huggingface.co/transformers/v2.2.0/model_doc/bert.html)."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "BXYitPoE-cjH"
+ },
+ "source": [
+ "\n",
+ "\n",
+ "We'll be using [BertForSequenceClassification](https://huggingface.co/transformers/v2.2.0/model_doc/bert.html#bertforsequenceclassification). This is the normal BERT model with an added single linear layer on top for classification that we will use as a sentence classifier. As we feed input data, the entire pre-trained BERT model and the additional untrained classification layer is trained on our specific task. \n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "WnQW9E-bBCRt"
+ },
+ "source": [
+ "OK, let's load BERT! There are a few different pre-trained BERT models available. \"bert-base-multilingual-cased\" means BERT base architecture with cased vocabulary for 104 languages. \"bert-base-multilingual-uncased\" means BERT base architecture with uncased vocabulary for 102 languages.\n",
+ "\n",
+ "The documentation for `from_pretrained` can be found [here](https://huggingface.co/transformers/pretrained_models.html), with the additional parameters defined [here](https://huggingface.co/transformers/v2.2.0/main_classes/configuration.html#transformers.PretrainedConfig)."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "gFsCTp_mporB",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000,
+ "referenced_widgets": [
+ "d3ac76358c3246f8897b55741031a3e5",
+ "38e2da0938504f90bd18c68d7e2ffc51",
+ "2e298550f61b4d08af0a6497d8e33b5b",
+ "62636381fe554245bb5efc22a509ffc1",
+ "5ccfbc88b71d4f15ba6bf374bb1007ed",
+ "a053a05834ae45838dca81a8ed4b0a27",
+ "474a715d1a054ec29547c1b54560f7f6",
+ "ed377908b70442429d666e123f526e0e",
+ "86290b6fa15844388475ac4a10d6a3aa",
+ "e77e34463f3f4de5abd1e8e9a28c4da7",
+ "0a1e169b0359416e93e9f2e5c6c7b1bb",
+ "7ad677239085495892d73c33ce07af73",
+ "fbf9d95bd65545cc8c51f3a5d80aef59",
+ "f31e5ef9468e4fefb41ed01d2ecb337c",
+ "084f12e8365d448fab0d70b7589182b7",
+ "c57df1509cd240f999d37e4a1bd5c075"
+ ]
+ },
+ "outputId": "a258a4eb-27bf-421e-f04b-d938799570b8"
+ },
+ "source": [
+ "from transformers import BertForSequenceClassification, AdamW, BertConfig\n",
+ "\n",
+ "# Load BertForSequenceClassification, the pretrained BERT model with a single \n",
+ "# linear classification layer on top. \n",
+ "model = BertForSequenceClassification.from_pretrained(\n",
+ " \"bert-base-multilingual-cased\", # Use the 12-layer BERT model, with an uncased vocab.\n",
+ " num_labels = 2, # The number of output labels--2 for binary classification.\n",
+ " # You can increase this for multi-class tasks. \n",
+ " output_attentions = False, # Whether the model returns attentions weights.\n",
+ " output_hidden_states = False, # Whether the model returns all hidden-states.\n",
+ ")\n",
+ "\n",
+ "# Tell pytorch to run this model on the GPU.\n",
+ "model.cuda()"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "d3ac76358c3246f8897b55741031a3e5",
+ "version_minor": 0,
+ "version_major": 2
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=625.0, style=ProgressStyle(description_…"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "86290b6fa15844388475ac4a10d6a3aa",
+ "version_minor": 0,
+ "version_major": 2
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=714314041.0, style=ProgressStyle(descri…"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "BertForSequenceClassification(\n",
+ " (bert): BertModel(\n",
+ " (embeddings): BertEmbeddings(\n",
+ " (word_embeddings): Embedding(119547, 768, padding_idx=0)\n",
+ " (position_embeddings): Embedding(512, 768)\n",
+ " (token_type_embeddings): Embedding(2, 768)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (encoder): BertEncoder(\n",
+ " (layer): ModuleList(\n",
+ " (0): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (1): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (2): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (3): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (4): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (5): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (6): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (7): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (8): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (9): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (10): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (11): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " )\n",
+ " )\n",
+ " (pooler): BertPooler(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (activation): Tanh()\n",
+ " )\n",
+ " )\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " (classifier): Linear(in_features=768, out_features=2, bias=True)\n",
+ ")"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 18
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "e0Jv6c7-HHDW"
+ },
+ "source": [
+ "We can browse all of the model's parameters by name here.\n",
+ "\n",
+ "In the below cell, I've printed out the names and dimensions of the weights for:\n",
+ "\n",
+ "1. The embedding layer.\n",
+ "2. The first of the twelve transformers.\n",
+ "3. The output layer.\n",
+ "\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "8PIiVlDYCtSq",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 671
+ },
+ "outputId": "e0b9cb7c-63b7-480e-dcb4-b968b25b7741"
+ },
+ "source": [
+ "# Get all of the model's parameters as a list of tuples.\n",
+ "params = list(model.named_parameters())\n",
+ "\n",
+ "print('The BERT model has {:} different named parameters.\\n'.format(len(params)))\n",
+ "\n",
+ "print('==== Embedding Layer ====\\n')\n",
+ "\n",
+ "for p in params[0:5]:\n",
+ " print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))\n",
+ "\n",
+ "print('\\n==== First Transformer ====\\n')\n",
+ "\n",
+ "for p in params[5:21]:\n",
+ " print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))\n",
+ "\n",
+ "print('\\n==== Output Layer ====\\n')\n",
+ "\n",
+ "for p in params[-4:]:\n",
+ " print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "The BERT model has 201 different named parameters.\n",
+ "\n",
+ "==== Embedding Layer ====\n",
+ "\n",
+ "bert.embeddings.word_embeddings.weight (119547, 768)\n",
+ "bert.embeddings.position_embeddings.weight (512, 768)\n",
+ "bert.embeddings.token_type_embeddings.weight (2, 768)\n",
+ "bert.embeddings.LayerNorm.weight (768,)\n",
+ "bert.embeddings.LayerNorm.bias (768,)\n",
+ "\n",
+ "==== First Transformer ====\n",
+ "\n",
+ "bert.encoder.layer.0.attention.self.query.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.self.query.bias (768,)\n",
+ "bert.encoder.layer.0.attention.self.key.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.self.key.bias (768,)\n",
+ "bert.encoder.layer.0.attention.self.value.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.self.value.bias (768,)\n",
+ "bert.encoder.layer.0.attention.output.dense.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.output.dense.bias (768,)\n",
+ "bert.encoder.layer.0.attention.output.LayerNorm.weight (768,)\n",
+ "bert.encoder.layer.0.attention.output.LayerNorm.bias (768,)\n",
+ "bert.encoder.layer.0.intermediate.dense.weight (3072, 768)\n",
+ "bert.encoder.layer.0.intermediate.dense.bias (3072,)\n",
+ "bert.encoder.layer.0.output.dense.weight (768, 3072)\n",
+ "bert.encoder.layer.0.output.dense.bias (768,)\n",
+ "bert.encoder.layer.0.output.LayerNorm.weight (768,)\n",
+ "bert.encoder.layer.0.output.LayerNorm.bias (768,)\n",
+ "\n",
+ "==== Output Layer ====\n",
+ "\n",
+ "bert.pooler.dense.weight (768, 768)\n",
+ "bert.pooler.dense.bias (768,)\n",
+ "classifier.weight (2, 768)\n",
+ "classifier.bias (2,)\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "qRWT-D4U_Pvx"
+ },
+ "source": [
+ "## 4.2. Optimizer & Learning Rate Scheduler"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "8o-VEBobKwHk"
+ },
+ "source": [
+ "Now that I have my model loaded I need to grab the training hyperparameters from within the stored model.\n",
+ "\n",
+ "For the purposes of fine-tuning, the authors recommend choosing from the following values:\n",
+ "- Batch size: 16, 32 (I chose 32 when creating our DataLoaders).\n",
+ "- Learning rate (Adam): 5e-5, 3e-5, 2e-5 (I'll use 2e-5).\n",
+ "- Number of epochs: 2, 3, 4 (I'll use 4).\n",
+ "\n",
+ "The epsilon parameter `eps = 1e-8` is \"a very small number to prevent any division by zero in the implementation\" (from [here](https://machinelearningmastery.com/adam-optimization-algorithm-for-deep-learning/)).\n",
+ "\n",
+ "You can find the creation of the AdamW optimizer in `run_glue.py` [here](https://github.com/huggingface/transformers/blob/5bfcd0485ece086ebcbed2d008813037968a9e58/examples/run_glue.py#L109)."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "GLs72DuMODJO"
+ },
+ "source": [
+ "# Note: AdamW is a class from the huggingface library (as opposed to pytorch) \n",
+ "# I believe the 'W' stands for 'Weight Decay fix\"\n",
+ "optimizer = AdamW(model.parameters(),\n",
+ " lr = 2e-5, # args.learning_rate - default is 5e-5, our notebook had 2e-5\n",
+ " eps = 1e-8 # args.adam_epsilon - default is 1e-8.\n",
+ " )"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "-p0upAhhRiIx"
+ },
+ "source": [
+ "from transformers import get_linear_schedule_with_warmup\n",
+ "\n",
+ "# Number of training epochs (authors recommend between 2 and 4)\n",
+ "epochs = 4\n",
+ "\n",
+ "# Total number of training steps is number of batches * number of epochs.\n",
+ "total_steps = len(train_dataloader) * epochs\n",
+ "\n",
+ "# Create the learning rate scheduler.\n",
+ "scheduler = get_linear_schedule_with_warmup(optimizer, \n",
+ " num_warmup_steps = 0, # Default value in run_glue.py\n",
+ " num_training_steps = total_steps)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "RqfmWwUR_Sox"
+ },
+ "source": [
+ "## 4.3. Training Loop"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "_QXZhFb4LnV5"
+ },
+ "source": [
+ "Below is the training loop. There's a lot going on, but fundamentally for each pass in this loop I have a trianing phase and a validation phase. At each pass I need to:\n",
+ "\n",
+ "Training loop:\n",
+ "- Unpack data inputs and labels\n",
+ "- Load data onto the GPU for acceleration\n",
+ "- Clear out the gradients calculated in the previous pass. \n",
+ " - In pytorch the gradients accumulate by default (useful for things like RNNs) unless you explicitly clear them out.\n",
+ "- Forward pass (feed input data through the network)\n",
+ "- Backward pass (backpropagation)\n",
+ "- Tell the network to update parameters with optimizer.step()\n",
+ "- Track variables for monitoring progress\n",
+ "\n",
+ "Evalution loop:\n",
+ "- Unpack data inputs and labels\n",
+ "- Load data onto the GPU for acceleration\n",
+ "- Forward pass (feed input data through the network)\n",
+ "- Compute loss on our validation data and track variables for monitoring progress\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "pE5B99H5H2-W"
+ },
+ "source": [
+ "Define a helper function for calculating accuracy."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "9cQNvaZ9bnyy"
+ },
+ "source": [
+ "import numpy as np\n",
+ "\n",
+ "# Function to calculate the accuracy of our predictions vs labels\n",
+ "def flat_accuracy(preds, labels):\n",
+ " pred_flat = np.argmax(preds, axis=1).flatten()\n",
+ " labels_flat = labels.flatten()\n",
+ " return np.sum(pred_flat == labels_flat) / len(labels_flat)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "KNhRtWPXH9C3"
+ },
+ "source": [
+ "Helper function for formatting elapsed times.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "gpt6tR83keZD"
+ },
+ "source": [
+ "import time\n",
+ "import datetime\n",
+ "\n",
+ "def format_time(elapsed):\n",
+ " '''\n",
+ " Takes a time in seconds and returns a string hh:mm:ss\n",
+ " '''\n",
+ " # Round to the nearest second.\n",
+ " elapsed_rounded = int(round((elapsed)))\n",
+ " \n",
+ " # Format as hh:mm:ss\n",
+ " return str(datetime.timedelta(seconds=elapsed_rounded))\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "cfNIhN19te3N"
+ },
+ "source": [
+ "The training loop"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "6J-FYdx6nFE_",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 876
+ },
+ "outputId": "090fe3e2-2c48-4f72-f44f-7d9172b17e05"
+ },
+ "source": [
+ "import random\n",
+ "\n",
+ "# This training code is based on the `run_glue.py` script here:\n",
+ "# https://github.com/huggingface/transformers/blob/5bfcd0485ece086ebcbed2d008813037968a9e58/examples/run_glue.py#L128\n",
+ "\n",
+ "\n",
+ "# Set the seed value all over the place to make this reproducible.\n",
+ "seed_val = 42\n",
+ "\n",
+ "random.seed(seed_val)\n",
+ "np.random.seed(seed_val)\n",
+ "torch.manual_seed(seed_val)\n",
+ "torch.cuda.manual_seed_all(seed_val)\n",
+ "\n",
+ "# Store the average loss after each epoch so I can plot them.\n",
+ "loss_values = []\n",
+ "\n",
+ "# For each epoch...\n",
+ "for epoch_i in range(0, epochs):\n",
+ " \n",
+ " # ========================================\n",
+ " # Training\n",
+ " # ========================================\n",
+ " \n",
+ " # Perform one full pass over the training set.\n",
+ "\n",
+ " print(\"\")\n",
+ " print('======== Epoch {:} / {:} ========'.format(epoch_i + 1, epochs))\n",
+ " print('Training...')\n",
+ "\n",
+ " # Measure how long the training epoch takes.\n",
+ " t0 = time.time()\n",
+ "\n",
+ " # Reset the total loss for this epoch.\n",
+ " total_loss = 0\n",
+ "\n",
+ " # Put the model into training mode.\n",
+ " model.train()\n",
+ "\n",
+ " # For each batch of training data...\n",
+ " for step, batch in enumerate(train_dataloader):\n",
+ "\n",
+ " # Progress update every 40 batches.\n",
+ " if step % 40 == 0 and not step == 0:\n",
+ " # Calculate elapsed time in minutes.\n",
+ " elapsed = format_time(time.time() - t0)\n",
+ " \n",
+ " # Report progress.\n",
+ " print(' Batch {:>5,} of {:>5,}. Elapsed: {:}.'.format(step, len(train_dataloader), elapsed))\n",
+ "\n",
+ " # Unpack this training batch from the dataloader. \n",
+ " #\n",
+ " # As I unpack the batch, I'll also copy each tensor to the GPU using the \n",
+ " # `to` method.\n",
+ " #\n",
+ " # `batch` contains three pytorch tensors:\n",
+ " # [0]: input ids \n",
+ " # [1]: attention masks\n",
+ " # [2]: labels \n",
+ " b_input_ids = batch[0].to(device)\n",
+ " b_input_mask = batch[1].to(device)\n",
+ " b_labels = batch[2].to(device)\n",
+ "\n",
+ " # Always clear any previously calculated gradients before performing a\n",
+ " # backward pass. PyTorch doesn't do this automatically because \n",
+ " # accumulating the gradients is \"convenient while training RNNs\". \n",
+ " # (source: https://stackoverflow.com/questions/48001598/why-do-we-need-to-call-zero-grad-in-pytorch)\n",
+ " model.zero_grad() \n",
+ "\n",
+ " # Perform a forward pass (evaluate the model on this training batch).\n",
+ " # This will return the loss (rather than the model output) because I\n",
+ " # have provided the `labels`.\n",
+ " # The documentation for this `model` function is here: \n",
+ " # https://huggingface.co/transformers/v2.2.0/model_doc/bert.html#transformers.BertForSequenceClassification\n",
+ " outputs = model(b_input_ids, \n",
+ " token_type_ids=None, \n",
+ " attention_mask=b_input_mask, \n",
+ " labels=b_labels)\n",
+ " \n",
+ " # The call to `model` always returns a tuple, so I need to pull the \n",
+ " # loss value out of the tuple.\n",
+ " loss = outputs[0]\n",
+ "\n",
+ " # Accumulate the training loss over all of the batches so that I can\n",
+ " # calculate the average loss at the end. `loss` is a Tensor containing a\n",
+ " # single value; the `.item()` function just returns the Python value \n",
+ " # from the tensor.\n",
+ " total_loss += loss.item()\n",
+ "\n",
+ " # Perform a backward pass to calculate the gradients.\n",
+ " loss.backward()\n",
+ "\n",
+ " # Clip the norm of the gradients to 1.0.\n",
+ " # This is to help prevent the \"exploding gradients\" problem.\n",
+ " torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0)\n",
+ "\n",
+ " # Update parameters and take a step using the computed gradient.\n",
+ " # The optimizer dictates the \"update rule\"--how the parameters are\n",
+ " # modified based on their gradients, the learning rate, etc.\n",
+ " optimizer.step()\n",
+ "\n",
+ " # Update the learning rate.\n",
+ " scheduler.step()\n",
+ "\n",
+ " # Calculate the average loss over the training data.\n",
+ " avg_train_loss = total_loss / len(train_dataloader) \n",
+ " \n",
+ " # Store the loss value for plotting the learning curve.\n",
+ " loss_values.append(avg_train_loss)\n",
+ "\n",
+ " print(\"\")\n",
+ " print(\" Average training loss: {0:.2f}\".format(avg_train_loss))\n",
+ " print(\" Training epcoh took: {:}\".format(format_time(time.time() - t0)))\n",
+ " \n",
+ " # ========================================\n",
+ " # Validation\n",
+ " # ========================================\n",
+ " # After the completion of each training epoch, measure the performance on\n",
+ " # the validation set.\n",
+ "\n",
+ " print(\"\")\n",
+ " print(\"Running Validation...\")\n",
+ "\n",
+ " t0 = time.time()\n",
+ "\n",
+ " # Put the model in evaluation mode--the dropout layers behave differently\n",
+ " # during evaluation.\n",
+ " model.eval()\n",
+ "\n",
+ " # Tracking variables \n",
+ " eval_loss, eval_accuracy = 0, 0\n",
+ " nb_eval_steps, nb_eval_examples = 0, 0\n",
+ "\n",
+ " # Evaluate data for one epoch\n",
+ " for batch in validation_dataloader:\n",
+ " \n",
+ " # Add batch to GPU\n",
+ " batch = tuple(t.to(device) for t in batch)\n",
+ " \n",
+ " # Unpack the inputs from dataloader\n",
+ " b_input_ids, b_input_mask, b_labels = batch\n",
+ " \n",
+ " # Telling the model not to compute or store gradients, saving memory and\n",
+ " # speeding up validation\n",
+ " with torch.no_grad(): \n",
+ "\n",
+ " # Forward pass, calculate logit predictions.\n",
+ " # This will return the logits rather than the loss because we have\n",
+ " # not provided labels.\n",
+ " # token_type_ids is the same as the \"segment ids\", which \n",
+ " # differentiates sentence 1 and 2 in 2-sentence tasks.\n",
+ " # The documentation for this `model` function is here: \n",
+ " # https://huggingface.co/transformers/v2.2.0/model_doc/bert.html#transformers.BertForSequenceClassification\n",
+ " outputs = model(b_input_ids, \n",
+ " token_type_ids=None, \n",
+ " attention_mask=b_input_mask)\n",
+ " \n",
+ " # Get the \"logits\" output by the model. The \"logits\" are the output\n",
+ " # values prior to applying an activation function like the softmax.\n",
+ " logits = outputs[0]\n",
+ "\n",
+ " # Move logits and labels to CPU\n",
+ " logits = logits.detach().cpu().numpy()\n",
+ " label_ids = b_labels.to('cpu').numpy()\n",
+ " \n",
+ " # Calculate the accuracy for this batch of test sentences.\n",
+ " tmp_eval_accuracy = flat_accuracy(logits, label_ids)\n",
+ " \n",
+ " # Accumulate the total accuracy.\n",
+ " eval_accuracy += tmp_eval_accuracy\n",
+ "\n",
+ " # Track the number of batches\n",
+ " nb_eval_steps += 1\n",
+ "\n",
+ " # Report the final accuracy for this validation run.\n",
+ " print(\" Accuracy: {0:.2f}\".format(eval_accuracy/nb_eval_steps))\n",
+ " print(\" Validation took: {:}\".format(format_time(time.time() - t0)))\n",
+ "\n",
+ "print(\"\")\n",
+ "print(\"Training complete!\")"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "\n",
+ "======== Epoch 1 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:51.\n",
+ "\n",
+ " Average training loss: 0.09\n",
+ " Training epcoh took: 0:01:24\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:03\n",
+ "\n",
+ "======== Epoch 2 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:53.\n",
+ "\n",
+ " Average training loss: 0.01\n",
+ " Training epcoh took: 0:01:27\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:04\n",
+ "\n",
+ "======== Epoch 3 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:56.\n",
+ "\n",
+ " Average training loss: 0.01\n",
+ " Training epcoh took: 0:01:31\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:04\n",
+ "\n",
+ "======== Epoch 4 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:56.\n",
+ "\n",
+ " Average training loss: 0.00\n",
+ " Training epcoh took: 0:01:32\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:04\n",
+ "\n",
+ "Training complete!\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "1-G03mmwH3aI"
+ },
+ "source": [
+ "Plot the training loss over all batches:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "68xreA9JAmG5",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 484
+ },
+ "outputId": "c5d61007-f966-4472-a1f0-0dd4ff7b8f4d"
+ },
+ "source": [
+ "import matplotlib.pyplot as plt\n",
+ "% matplotlib inline\n",
+ "\n",
+ "import seaborn as sns\n",
+ "\n",
+ "# Use plot styling from seaborn.\n",
+ "sns.set(style='darkgrid')\n",
+ "\n",
+ "# Increase the plot size and font size.\n",
+ "sns.set(font_scale=1.5)\n",
+ "plt.rcParams[\"figure.figsize\"] = (12,6)\n",
+ "\n",
+ "# Plot the learning curve.\n",
+ "plt.plot(loss_values, 'b-o')\n",
+ "\n",
+ "# Label the plot.\n",
+ "plt.title(\"Training loss\")\n",
+ "plt.xlabel(\"Epoch\")\n",
+ "plt.ylabel(\"Loss\")\n",
+ "\n",
+ "plt.show()"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
+ " import pandas.util.testing as tm\n"
+ ],
+ "name": "stderr"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 864x432 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "mkyubuJSOzg3"
+ },
+ "source": [
+ "# 5. Performance On Test Set"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "DosV94BYIYxg"
+ },
+ "source": [
+ "Now I'll load the holdout dataset and prepare inputs just as I did with the training set. Then I'll evaluate predictions using many matrics like ( F1_score, Precision, Recall, accuracy...)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Tg42jJqqM68F"
+ },
+ "source": [
+ "### 5.1. Data Preparation\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "xWe0_JW21MyV"
+ },
+ "source": [
+ "I'll need to apply all of the same steps that I did for the training data to prepare the test data set."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "mAN0LZBOOPVh",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "bc6e4f0e-5fcd-4442-ccd7-c1e7992e311d"
+ },
+ "source": [
+ "import pandas as pd\n",
+ "\n",
+ "# Load the dataset into a pandas dataframe.\n",
+ "df = pd.read_csv(\"/content/test_fin_07_06.csv\")\n",
+ "\n",
+ "# Report the number of sentences.\n",
+ "print('Number of test sentences: {:,}\\n'.format(df.shape[0]))\n",
+ "\n",
+ "# Create sentence and label lists\n",
+ "sentences = df.sentences.values\n",
+ "labels = df.labels.values\n",
+ "\n",
+ "\n"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Number of test sentences: 2,412\n",
+ "\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "3jGG2VzTgF2z",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 73
+ },
+ "outputId": "b7b5f4f2-b5e9-45ad-dbbf-a7944a948410"
+ },
+ "source": [
+ "df['labels'].value_counts()"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0 1871\n",
+ "1 541\n",
+ "Name: labels, dtype: int64"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 51
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "bzzq0wGeIHUB"
+ },
+ "source": [
+ "# Tokenize all of the sentences and map the tokens to thier word IDs.\n",
+ "input_ids = []\n",
+ "# For every sentence...\n",
+ "for sent in sentences:\n",
+ " # `encode` will:\n",
+ " # (1) Tokenize the sentence.\n",
+ " # (2) Prepend the `[CLS]` token to the start.\n",
+ " # (3) Append the `[SEP]` token to the end.\n",
+ " # (4) Map tokens to their IDs.\n",
+ " encoded_sent = tokenizer.encode(\n",
+ " sent, # Sentence to encode.\n",
+ " add_special_tokens = True, # Add '[CLS]' and '[SEP]'\n",
+ " )\n",
+ " \n",
+ " input_ids.append(encoded_sent)\n",
+ "\n",
+ "# Pad our input tokens\n",
+ "input_ids = pad_sequences(input_ids, maxlen=MAX_LEN, \n",
+ " dtype=\"long\", truncating=\"post\", padding=\"post\")\n",
+ "\n",
+ "# Create attention masks\n",
+ "attention_masks = []\n",
+ "\n",
+ "# Create a mask of 1s for each token followed by 0s for padding\n",
+ "for seq in input_ids:\n",
+ " seq_mask = [float(i>0) for i in seq]\n",
+ " attention_masks.append(seq_mask) \n",
+ "\n",
+ "# Convert to tensors.\n",
+ "prediction_inputs = torch.tensor(input_ids)\n",
+ "prediction_masks = torch.tensor(attention_masks)\n",
+ "prediction_labels = torch.tensor(labels)\n",
+ "\n",
+ "# Set the batch size. \n",
+ "batch_size = 32 \n",
+ "\n",
+ "# Create the DataLoader.\n",
+ "prediction_data = TensorDataset(prediction_inputs, prediction_masks, prediction_labels)\n",
+ "prediction_sampler = SequentialSampler(prediction_data)\n",
+ "prediction_dataloader = DataLoader(prediction_data, sampler=prediction_sampler, batch_size=batch_size)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "16lctEOyNFik"
+ },
+ "source": [
+ "## 5.2. Evaluate on Test Set\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "rhR99IISNMg9"
+ },
+ "source": [
+ "\n",
+ "With the test set prepared, I can apply the fine-tuned model to generate predictions on the test set."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Hba10sXR7Xi6",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "6f7abf0e-87db-4a7e-b6cc-8dedbc31f795"
+ },
+ "source": [
+ "# Prediction on test set\n",
+ "\n",
+ "print('Predicting labels for {:,} test sentences...'.format(len(prediction_inputs)))\n",
+ "\n",
+ "# Put model in evaluation mode\n",
+ "model.eval()\n",
+ "\n",
+ "# Tracking variables \n",
+ "predictions_test , true_labels = [], []\n",
+ "\n",
+ "# Predict \n",
+ "for batch in prediction_dataloader:\n",
+ "# Add batch to GPU\n",
+ " batch = tuple(t.to(device) for t in batch)\n",
+ " \n",
+ " # Unpack the inputs from the dataloader\n",
+ " b_input_ids, b_input_mask, b_labels = batch\n",
+ " \n",
+ " # Telling the model not to compute or store gradients, saving memory and \n",
+ " # speeding up prediction\n",
+ " with torch.no_grad():\n",
+ " # Forward pass, calculate logit predictions\n",
+ " outputs = model(b_input_ids, token_type_ids=None, \n",
+ " attention_mask=b_input_mask)\n",
+ "\n",
+ " logits = outputs[0]\n",
+ " #print(logits)\n",
+ "\n",
+ " # Move logits and labels to CPU\n",
+ " logits = logits.detach().cpu().numpy()\n",
+ " label_ids = b_labels.to('cpu').numpy()\n",
+ " #print(logits)\n",
+ " \n",
+ " # Store predictions and true labels\n",
+ " predictions_test.append(logits)\n",
+ " true_labels.append(label_ids)\n",
+ "\n",
+ "print(' DONE.')"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Predicting labels for 2,412 test sentences...\n",
+ " DONE.\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "LtmGGxm8EJjX",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "0ec619f6-b549-410e-c272-17c2e0a01df9"
+ },
+ "source": [
+ "predictions_test[0][0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([ 4.3997846, -3.7260323], dtype=float32)"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 54
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "hWcy0X1hirdx",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "7df9fb9c-58c7-49fd-debf-21856c034674"
+ },
+ "source": [
+ "print('Positive samples: %d of %d (%.2f%%)' % (df.labels.sum(), len(df.labels), (df.labels.sum() / len(df.labels) * 100.0)))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Positive samples: 541 of 2412 (22.43%)\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "-5jscIM8R4Gv"
+ },
+ "source": [
+ "We use MCC here because the classes are imbalanced:\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "cRaZQ4XC7kLs",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "5528b943-47f9-46e4-a891-5336e88a0e47"
+ },
+ "source": [
+ "from sklearn.metrics import *\n",
+ "\n",
+ "matthews_set = []\n",
+ "pred_labels = []\n",
+ "f1_score_met = []\n",
+ "roc_auc_score_met = []\n",
+ "precision_score_met = []\n",
+ "recall_score_met = []\n",
+ "confusion_matrix_met =[]\n",
+ "precision_recall_curve_met =[]\n",
+ "accuracy_score_met = []\n",
+ "# Evaluate each test batch using many matrics\n",
+ "print('Calculating the matrics for each batch...')\n",
+ "\n",
+ "# For each input batch...\n",
+ "for i in range(len(true_labels)):\n",
+ " \n",
+ " # The predictions for this batch are a 2-column ndarray (one column for \"0\" \n",
+ " # and one column for \"1\"). Pick the label with the highest value and turn this\n",
+ " # in to a list of 0s and 1s.\n",
+ " pred_labels_i = np.argmax(predictions_test[i], axis=1).flatten()\n",
+ " pred_labels.append(pred_labels_i)\n",
+ "\n",
+ "# Calculate and store the similarity for this batch depending on accuracy\n",
+ " accuracy_score_i = accuracy_score(pred_labels_i, true_labels[i])\n",
+ " accuracy_score_met.append(accuracy_score_i)\n",
+ "\n",
+ "# Calculate and store the similarity for this batch depending on F1_score\n",
+ " f1_score_i = f1_score(pred_labels_i, true_labels[i], average=\"binary\")\n",
+ " f1_score_met.append(f1_score_i)\n",
+ "\n",
+ "\n",
+ " confusion_matrix_i = confusion_matrix(pred_labels_i, true_labels[i])\n",
+ " confusion_matrix_met.append(confusion_matrix_i)\n",
+ "\n",
+ "\n",
+ "# Calculate and store the similarity for this batch depending on precision, recall, and roc_auc_score\n",
+ " precision_recall_curve_i = precision_recall_curve(pred_labels_i, true_labels[i])\n",
+ " precision_recall_curve_met.append(precision_recall_curve_i)\n",
+ "\n",
+ " precision_score_i = precision_score(pred_labels_i, true_labels[i], average=\"binary\")\n",
+ " precision_score_met.append(precision_score_i)\n",
+ "\n",
+ " roc_auc_score_i = roc_auc_score(pred_labels_i, true_labels[i])\n",
+ " roc_auc_score_met.append(roc_auc_score_i)\n",
+ "\n",
+ " recall_score_i = recall_score(pred_labels_i, true_labels[i], average=\"binary\")\n",
+ " recall_score_met.append(recall_score_i)\n",
+ "\n",
+ " # Calculate and store the coef for this batch depending on mathew\n",
+ " matthews = matthews_corrcoef(true_labels[i], pred_labels_i) \n",
+ " matthews_set.append(matthews)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Calculating Matthews Corr. Coef. for each batch...\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "FmYcYZKF7ooS"
+ },
+ "source": [
+ "#Calculate the number of Tp,Tn,Fp and Fn in the predictions\n",
+ "tns = 0\n",
+ "fps = 0\n",
+ "fns = 0\n",
+ "tps = 0\n",
+ "for i in range(len(predictions_test)):\n",
+ " pred_labels_i_fin = np.argmax(predictions_test[i], axis=1).flatten()\n",
+ " tn, fp, fn, tp = confusion_matrix(pred_labels_i_fin, true_labels[i]).ravel()\n",
+ " tns = tns + tn\n",
+ " fps = fps + fp\n",
+ " fns = fns + fn\n",
+ " tps = tps + tp"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "s07p1BIzrmFq",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "360324aa-113b-43bc-bc62-e232844c3ba3"
+ },
+ "source": [
+ "#Calculate the number of Tp,Tn,Fp and Fn in the predictions by confusion matrix\n",
+ "sum(confusion_matrix_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([[1797, 0],\n",
+ " [ 74, 541]])"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 58
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "zhyB219mwQwg",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "0b366c6b-b563-4216-d08e-78524ec960fc"
+ },
+ "source": [
+ "#Sample of true labels in the dataset\n",
+ "true_labels[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0,\n",
+ " 0, 1, 0, 0, 0, 0, 0, 0, 1, 0])"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 59
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Iyn-reIKtEG8",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "d0d745b2-183d-45da-ac8e-169a4759cfcd"
+ },
+ "source": [
+ "#Sample of predicted labels in the predictions list\n",
+ "pred_labels[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0,\n",
+ " 0, 1, 0, 0, 0, 0, 0, 0, 1, 0])"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 60
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "a1AWkA3POWBQ"
+ },
+ "source": [
+ "import seaborn as sn\n",
+ "import matplotlib.pyplot as plt"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "AdR1y2u2N5d6",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000
+ },
+ "outputId": "1780d9a7-ac8d-4128-d296-20cdb6aacca0"
+ },
+ "source": [
+ "#Plot the confusion matrix for each batch\n",
+ "for i in range(len(confusion_matrix_met)):\n",
+ " df_cm = pd.DataFrame(confusion_matrix_met[i], index = [i for i in \"FT\"],\n",
+ " columns = [i for i in \"FT\"])\n",
+ " plt.figure(figsize = (3,2))\n",
+ " sn.heatmap(df_cm, annot=True)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\n",
+ " \"\"\"\n"
+ ],
+ "name": "stderr"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARtUlEQVR4nO3deVQUV74H8G8D0my2YF6jiHFLlEUUROOoz6go4aBGkURpo6JEwQ1wRDM+MyZj9GTUhxg3MIm4kahoNCAax5goPk/yXKMRw4BmZFyGtEAzPG1l6f39waFDWy3dstzqon+fc+oc+96i6/dHf723bnVXiQwGgwGEECYc+C6AEHtCgSOEIQocIQxR4AhhiAJHCEMUOEIYcuK7gMY0lf/kuwSb4Nrtdb5LsCla9W9N9pv73HT4jz5tVU6L2FTgCGkWjYrvCqxGgSOCZ9Bp+S7BahQ4InxaGuEIYYZGOEJYosARwhAtmhDCEI1whLBj0Gv4LsFqFDgifDSlJIQhmlISwhAFjhB2DDSlJIQhGuEIYYgCRwhDGjXfFViNAkeEr4Uj3M2bN5Gbm4vLly9DLpfD09MTgwYNwtKlS9GzZ0+Tfa9fv46NGzeiqKgIHh4eGD9+PJYvXw5XV1erjkWBI8KnbdkIt2vXLly/fh2RkZHw8/ODQqHAgQMHMGXKFBw9ehSvvPIKAKC4uBhxcXF49dVXsXLlSpSVlWHPnj0oLS3FZ599ZtWxKHBE+LQtG+Hi4uKQlpYGZ2dnY9uECRMwadIkZGZmYsOGDQCATz75BJ6envjyyy/h7u4OAOjevTs++OADXLx4EcOHD7d4LLsJ3L0HpfjmdD4uXLmOf8kfQqXS4GVfH0SMHYnYmGi4uboY983YvR+f7jlg9n2WJ87DuzOmsiqbOZFIhCXJ8UhImIVePbtDoajC0aMnsHrNRtTU1PJdnnk6XYv+PDQ0lNPWq1cv9O3bFyUlJQCAp0+f4sKFC5g3b54xbAAQFRWFdevW4dSpUxS4xnJPfofsr79B2Mg/YGJEGJycnHDl+k1s3/kFTuf/gIM7N8NFLDb5m/9aMh+enp1M2gL9XmVZNnOb0j7CkuR45B77GzZv/hwB/n2RlDQXISFBiIiUwSbvjG9m0USpVEKpVHLaJRIJJBKJxbc0GAyorKyEv78/AOD27dvQarUICgoy2c/Z2RkBAQEoLi62qlS7CdwbY0YiPlaGjh6//+8ki56Ini93w86sQ8g5cRozpk42+Zuxo0bA16cL61J5ExjYD0mJc5GTexIxsvnG9rv3HmDrlo8hk0Xh0KFjPFb4HGZGuKysLKSnp3Pak5KSkJycbPEtjx8/jvLycqSkpAAAFAoFAEAqlXL2lUqluHHjhlWl2k3gggL6mW2PHDcKO7MO4R//vG+2/2l1NVzELnBycmzL8mzCdNkUODg4YNu2XSbtu3YfxLq//hkz33nLRgPHPYebM2cOoqOjOe3WjG4lJSVYu3YtBg8ejKioKABAXV0dAJic5zUQi8XGfkusClxmZibGjh1rXK1pT8orKgEAL3X25PS9NXsRqmtq4ejogKAAPyyMewevD3+NdYnMDBkcDJ1OhytXTf+3VqlUKCj4O4YMCeGpsqYZNNyf51g7dXyWQqHAggUL0KlTJ2zduhUODvW3bnVxqT/HV6u501eVSmXst8SqG8Fu2rQJRUVFxtePHj1CaGgorl69atVBbJVOp8Nn+7Lh5OiIiW+EGdslHu6YFjUef05ZhO0bVuOPC97Fw7IKLP7Tahw7+T2PFbctn25dUFlZZfZD9Zu8DFLpS+jQoQMPlVmg1XG3Znjy5AkSEhLw5MkT7Nq1y2T62PDvhqllYwqFAt7e3lYdo1lTSoPBgJqaGmhbuBzLt//e+jkKCovxxwVx6N2zu7E9VmY6FQkD8NabEZgSuxCp23ciImwk3Nysu9ApJG6urlCpzF/Tqqur/4Kwm5srHj+2sR98tnCVEqgfpRYuXIh79+5h37596NPH9Eay/fr1g5OTEwoLCxEREWFsV6vVKC4uxqRJk6w6jt3e6nz7zi9w8OsTmBY1HgmzZRb39+wkQcyUiVA+eYqffymyuL8Q1dTWQizmnqMAgItL/QquTV4a0Gi42wvQ6XRYunQpbty4ga1btyIkhDt17tixI4YPH468vDxUV1cb2/Py8lBTU4PIyEirjmU3iyaNZezej8+zsjFl4hv4y58sr1g18O1aP2149Ji73NwePJSXIzCgH5ydnTnTSt9uXaFQ/BuaF/wws2Bo5hSywYYNG5Cfn4+wsDA8evQIeXl5xj53d3eEh4cDAFJSUjB9+nTExsZi2rRpKCsrw969ezFq1CiMGDHCqmNZHbiHDx/i1q1bAOrnugBQWlpqbHtWw/ULW9NwUTtqfDjWrlwKkUhk9d/eL5UDML/A0h78dK0AERFjMPS1EPz4v1eM7WKxGMHB/fHDD5d4rK4JLZxSNnyGz507h3Pnzpn0+fr6GgPXv39/7N27F2lpaVi/fj08PDwQExODZcuWWX0skTXP+Pb39+d8MA0Gg9kPa0O7tRcCG2vrh3l8uucAMnbvx6TIcfjrqmXGFajGtFodauvqTK7XAcDDcgWmxiVCJBLhTO4XnIvkrYmvh3kEBfnj+k/f41jeKZPrcImL38XWLR9jdlwyDh7MYV6XpYd5VP9lOqfNfe2htiqnRawa4davX9/WdbS57K9PIGP3fvh08cawISE4+f3/mPS/5OWJEUNDUVNbi8hp72Ls68PRp9fLkHT0wN0Hpcg5cRo1tbVI/Whlm4aNT4WFt7Dj031ISpyLI19l4tSpfOM3Tc6fv4Ds7Fy+SzSvFRZNWLEqcOYuIApNYfGvAICH5RVY9fEmTv+QQQMwYmgoXMTOCB/9n/il6Dbyf7iImppaeHpKMGxICObOnIYBgX6sS2dq2fLVuH+/FPHxMzFh/DhUVlYhI2MvVq/ZaJtf60LLz+FYsmpKyQo9H64ePR/OlKUp5dNlkzltHp8cb6tyWsQuVylJ+2LQ6vkuwWoUOCJ8auF8AYMCRwTPoKMRjhBmaEpJCEMGNQWOEHa0NrPQbhEFjgiegQJHCDt6NQWOEGYMwrkqQIEjwkeBI4Qhvcb6n1jxjQJHBE+vpcARwoxeR4EjhBkdTSkJYUevFc69sChwRPB0FDhC2NHRORwh7Oh1NMIRwoyWppSEsKPX05SyWejmOfUWdxvJdwmCotPTCEcIM1o6hyOEHZ2BppSEMENTSkIY0tAIRwg7OgE95lA4lRLyHDqIONuLqqioQFpaGmJjYzFo0CD4+fnh8uXLZvc9e/YsoqOjMWDAAIwZMwbp6elWPw2YAkcETwMRZ3tRd+/eRWZmJsrLy+Hn9/wHtpw/fx6JiYno1KkTPvzwQ4SHhyMjI8PqJ0zRlJIInvYFHqr5PP3798elS5fg5eWFM2fOIDEx0ex+qampCAwMxO7du+Ho6Aig/impO3fuRGxsLHr16tXkcWiEI4KnM7O9KA8PD3h5eTW5z507d3Dnzh3IZDJj2ABgxowZ0Ov1+O677yweh0Y4IniaVhjhrFFUVAQACAoKMmnv0qULunbtauxvCgWOCJ65KaVSqYRSqeS0SyQSSCSSZh1HoVAAAKRSKadPKpWioqLC4ntQ4IjgmbvDQlZWFtLT0zntSUlJSE5ObtZx6urqAADOzs6cPrFYjNraWovvQYEjgmfu96dz5swx+6js5o5uAODi4gIAUKvVnD6VSmXsbwoFjgieuStgLZk6Pk/DVFKhUMDb29ukT6FQYNCgQRbfg1YpieBpRNytLQQEBAAACgsLTdrLy8tRVlZm7G8KBY4Ink7E3dpC37590adPHxw+fBg63e8XH7Kzs+Hg4ICIiAiL70FTSiJ4rfVogR07dgAASkpKAAB5eXm4du0aJBIJZs2aBQBYsWIFFi1ahHnz5mHChAn49ddfceDAAchkMvTu3dviMUQGg8FmnvXj5OzLdwk2gX7xbWrbvcNN9m/oOYvTtvL+/hc+zvO+0uXr64v8/Hzj6zNnziA9PR0lJSXo3Lkz3n77bSxevBhOTpbHLxrhiODp0Dpjxu3bt63aLzw8HOHh4c06BgWOCF5zvsrFFwocAJFIhCXJ8UhImIVePbtDoajC0aMnsHrNRtTUWL6Y2V6MXzoV45dOe26/TqNFSt+ZDCuyjlpkM2dFFjUZOLlcjs6dO1t1QU/INqV9hCXJ8cg99jds3vw5Avz7IilpLkJCghARKYMNnea2qYJvr0Bxr4zT3s2/J8IXTkbh2es8VGVZuxnhxo0bh9TUVEyaNIlVPcwFBvZDUuJc5OSeRIxsvrH97r0H2LrlY8hkUTh06BiPFbIjv/UA8lsPOO2ydfXXly4ezuf02YLWOodjocnrcPbwP/t02RQ4ODhg27ZdJu27dh9EdXUNZr7zFk+V2QZnVzFC3xyB/5NXovj8Db7LMUsDA2ezVXZ/4XvI4GDodDpcuWr6YVKpVCgo+DuGDAnhqTLbEDJxGFwlbrh89DwMetv8IOtg4Gy2ymLgRIx+a8QXn25dUFlZZfYLqb/JyyCVvoQOHTrwUJltGB4TBr1ej0tfneO7lOcS0ghncZVy3bp12Lx5s1VvJhKJcObMmRYXxZKbqytUKm7YAKCuTlW/j5srHj/WsCzLJnj38cErQwNw+8dfUFWq4Luc59LacMCeZTFwPj4+6Nq1K4taeFFTWwtvD3ezfS4u4vp97OjSQGPDYsYCsN3Fkga2PIV8lsXAxcXFtetVyofycgQG9IOzszNnWunbrSsUin9Do7G/0c3B0QFD3x6Fp1VK3Dx9he9ymqQx6PkuwWp2v2jy07UCODo6YuhrposjYrEYwcH9ce1aAU+V8SsofDAkUk/8dOxHaNWt9fXgttGuFk3au6+OHIder8eSJfEm7fHzZsDd3Q0HD+XyVBm/hsWEAbD96SQgrMDZ/Ve7CgtvYcen+5CUOBdHvsrEqVP5xm+anD9/AdnZ9hc4ibcXAkaH4N6Nf+Dh7X/xXY5FGghnStlk4G7dusWqDl4tW74a9++XIj5+JiaMH4fKyipkZOzF6jUb7eLi/7P+MHU0HJ0ccfGQ7Y9uAKAT0Dkc/R7OBtHv4UxZ+j3cmz0mctq+eXCyrcppEbufUhLhE9IqJQWOCJ6uvZzDESIEQjqHo8ARwdNS4AhhR0tTSkLYoSklIQxpDcK5yQIFjggejXCEMEQjHCEM0QhHCEMUOEIY0upt+/d6jVHgiODRCEcIQxQ4QhjS6oWzSmn3t1ggwqcz6Dnbi1Kr1di4cSNGjhyJgQMHIiYmBhcvXmz1WilwRPB0ej1ne1ErV65EVlYWJk+ejFWrVsHBwQEJCQn4+eefW7VWChwRPI1ex9lexM2bN3Hy5Em89957WLFiBWQyGbKysuDj44O0tLRWrZUCRwSvpVPKb7/9Fh06dMC0ab8/G08sFmPq1Km4du0aKioqWq1WWjQhgqc3EzClUgmlUslpl0gkkEgkJm3FxcXo3bs33N1N78A9cOBAGAwGFBcXw9vbu1VqtanAadW/8V0CESBVHfdWftu3b0d6ejqnPSkpCcnJySZtCoUCXbp04ewrlUoBgEY4QiyZM2cOoqOjOe3Pjm4AUFdXZ/YJSWJx/bMlVCpVq9VFgSPtkrmp4/O4uLiYfX5EQ9AagtcaaNGE2D2pVGp22qhQ1D+iq7XO3wAKHCHw9/fH3bt3UV1dbdJeUFBg7G8tFDhi9yIjI6HRaHDkyBFjm1qtRk5ODkJDQ80uqDQXncMRuxccHIzIyEikpaVBoVCgR48eyM3NhVwux/r161v1WDb1bAFC+KJSqbBlyxacOHECjx8/hp+fH5YtW4YRI0a06nEocIQwROdwhDBEgSOEIVo0AZCTk4P333/fbN/y5csxf/58xhWx5+fnZ9V+Z8+eRffu3du4mvaLAtdISkoKfHx8TNoCAwN5qoat1NRUk9dZWVmQy+Wc/4g6d+7Msqx2hwLXyOjRoxEQEMB3GbyIiooyeX369Gk8evSI005ahs7hCGGIRrhGlEolqqqqjK9FIhG8vLx4rIi0NxS4RmbPnm3y2s3NrdXvaUHsGwWukTVr1qBHjx7G146OjjxWQ9ojClwjwcHBdrtoQtigRRNCGKLAEcIQBY4QhihwhDBEgSOEIfo9HCEM0QhHCEMUOEIYosARwhAFjhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQxQ4Qhj6fzsSRNhU3tgvAAAAAElFTkSuQmCC\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARtUlEQVR4nO3deVQUV74H8G8D0my2YF6jiHFLlEUUROOoz6go4aBGkURpo6JEwQ1wRDM+MyZj9GTUhxg3MIm4kahoNCAax5goPk/yXKMRw4BmZFyGtEAzPG1l6f39waFDWy3dstzqon+fc+oc+96i6/dHf723bnVXiQwGgwGEECYc+C6AEHtCgSOEIQocIQxR4AhhiAJHCEMUOEIYcuK7gMY0lf/kuwSb4Nrtdb5LsCla9W9N9pv73HT4jz5tVU6L2FTgCGkWjYrvCqxGgSOCZ9Bp+S7BahQ4InxaGuEIYYZGOEJYosARwhAtmhDCEI1whLBj0Gv4LsFqFDgifDSlJIQhmlISwhAFjhB2DDSlJIQhGuEIYYgCRwhDGjXfFViNAkeEr4Uj3M2bN5Gbm4vLly9DLpfD09MTgwYNwtKlS9GzZ0+Tfa9fv46NGzeiqKgIHh4eGD9+PJYvXw5XV1erjkWBI8KnbdkIt2vXLly/fh2RkZHw8/ODQqHAgQMHMGXKFBw9ehSvvPIKAKC4uBhxcXF49dVXsXLlSpSVlWHPnj0oLS3FZ599ZtWxKHBE+LQtG+Hi4uKQlpYGZ2dnY9uECRMwadIkZGZmYsOGDQCATz75BJ6envjyyy/h7u4OAOjevTs++OADXLx4EcOHD7d4LLsJ3L0HpfjmdD4uXLmOf8kfQqXS4GVfH0SMHYnYmGi4uboY983YvR+f7jlg9n2WJ87DuzOmsiqbOZFIhCXJ8UhImIVePbtDoajC0aMnsHrNRtTU1PJdnnk6XYv+PDQ0lNPWq1cv9O3bFyUlJQCAp0+f4sKFC5g3b54xbAAQFRWFdevW4dSpUxS4xnJPfofsr79B2Mg/YGJEGJycnHDl+k1s3/kFTuf/gIM7N8NFLDb5m/9aMh+enp1M2gL9XmVZNnOb0j7CkuR45B77GzZv/hwB/n2RlDQXISFBiIiUwSbvjG9m0USpVEKpVHLaJRIJJBKJxbc0GAyorKyEv78/AOD27dvQarUICgoy2c/Z2RkBAQEoLi62qlS7CdwbY0YiPlaGjh6//+8ki56Ini93w86sQ8g5cRozpk42+Zuxo0bA16cL61J5ExjYD0mJc5GTexIxsvnG9rv3HmDrlo8hk0Xh0KFjPFb4HGZGuKysLKSnp3Pak5KSkJycbPEtjx8/jvLycqSkpAAAFAoFAEAqlXL2lUqluHHjhlWl2k3gggL6mW2PHDcKO7MO4R//vG+2/2l1NVzELnBycmzL8mzCdNkUODg4YNu2XSbtu3YfxLq//hkz33nLRgPHPYebM2cOoqOjOe3WjG4lJSVYu3YtBg8ejKioKABAXV0dAJic5zUQi8XGfkusClxmZibGjh1rXK1pT8orKgEAL3X25PS9NXsRqmtq4ejogKAAPyyMewevD3+NdYnMDBkcDJ1OhytXTf+3VqlUKCj4O4YMCeGpsqYZNNyf51g7dXyWQqHAggUL0KlTJ2zduhUODvW3bnVxqT/HV6u501eVSmXst8SqG8Fu2rQJRUVFxtePHj1CaGgorl69atVBbJVOp8Nn+7Lh5OiIiW+EGdslHu6YFjUef05ZhO0bVuOPC97Fw7IKLP7Tahw7+T2PFbctn25dUFlZZfZD9Zu8DFLpS+jQoQMPlVmg1XG3Znjy5AkSEhLw5MkT7Nq1y2T62PDvhqllYwqFAt7e3lYdo1lTSoPBgJqaGmhbuBzLt//e+jkKCovxxwVx6N2zu7E9VmY6FQkD8NabEZgSuxCp23ciImwk3Nysu9ApJG6urlCpzF/Tqqur/4Kwm5srHj+2sR98tnCVEqgfpRYuXIh79+5h37596NPH9Eay/fr1g5OTEwoLCxEREWFsV6vVKC4uxqRJk6w6jt3e6nz7zi9w8OsTmBY1HgmzZRb39+wkQcyUiVA+eYqffymyuL8Q1dTWQizmnqMAgItL/QquTV4a0Gi42wvQ6XRYunQpbty4ga1btyIkhDt17tixI4YPH468vDxUV1cb2/Py8lBTU4PIyEirjmU3iyaNZezej8+zsjFl4hv4y58sr1g18O1aP2149Ji73NwePJSXIzCgH5ydnTnTSt9uXaFQ/BuaF/wws2Bo5hSywYYNG5Cfn4+wsDA8evQIeXl5xj53d3eEh4cDAFJSUjB9+nTExsZi2rRpKCsrw969ezFq1CiMGDHCqmNZHbiHDx/i1q1bAOrnugBQWlpqbHtWw/ULW9NwUTtqfDjWrlwKkUhk9d/eL5UDML/A0h78dK0AERFjMPS1EPz4v1eM7WKxGMHB/fHDD5d4rK4JLZxSNnyGz507h3Pnzpn0+fr6GgPXv39/7N27F2lpaVi/fj08PDwQExODZcuWWX0skTXP+Pb39+d8MA0Gg9kPa0O7tRcCG2vrh3l8uucAMnbvx6TIcfjrqmXGFajGtFodauvqTK7XAcDDcgWmxiVCJBLhTO4XnIvkrYmvh3kEBfnj+k/f41jeKZPrcImL38XWLR9jdlwyDh7MYV6XpYd5VP9lOqfNfe2htiqnRawa4davX9/WdbS57K9PIGP3fvh08cawISE4+f3/mPS/5OWJEUNDUVNbi8hp72Ls68PRp9fLkHT0wN0Hpcg5cRo1tbVI/Whlm4aNT4WFt7Dj031ISpyLI19l4tSpfOM3Tc6fv4Ds7Fy+SzSvFRZNWLEqcOYuIApNYfGvAICH5RVY9fEmTv+QQQMwYmgoXMTOCB/9n/il6Dbyf7iImppaeHpKMGxICObOnIYBgX6sS2dq2fLVuH+/FPHxMzFh/DhUVlYhI2MvVq/ZaJtf60LLz+FYsmpKyQo9H64ePR/OlKUp5dNlkzltHp8cb6tyWsQuVylJ+2LQ6vkuwWoUOCJ8auF8AYMCRwTPoKMRjhBmaEpJCEMGNQWOEHa0NrPQbhEFjgiegQJHCDt6NQWOEGYMwrkqQIEjwkeBI4Qhvcb6n1jxjQJHBE+vpcARwoxeR4EjhBkdTSkJYUevFc69sChwRPB0FDhC2NHRORwh7Oh1NMIRwoyWppSEsKPX05SyWejmOfUWdxvJdwmCotPTCEcIM1o6hyOEHZ2BppSEMENTSkIY0tAIRwg7OgE95lA4lRLyHDqIONuLqqioQFpaGmJjYzFo0CD4+fnh8uXLZvc9e/YsoqOjMWDAAIwZMwbp6elWPw2YAkcETwMRZ3tRd+/eRWZmJsrLy+Hn9/wHtpw/fx6JiYno1KkTPvzwQ4SHhyMjI8PqJ0zRlJIInvYFHqr5PP3798elS5fg5eWFM2fOIDEx0ex+qampCAwMxO7du+Ho6Aig/impO3fuRGxsLHr16tXkcWiEI4KnM7O9KA8PD3h5eTW5z507d3Dnzh3IZDJj2ABgxowZ0Ov1+O677yweh0Y4IniaVhjhrFFUVAQACAoKMmnv0qULunbtauxvCgWOCJ65KaVSqYRSqeS0SyQSSCSSZh1HoVAAAKRSKadPKpWioqLC4ntQ4IjgmbvDQlZWFtLT0zntSUlJSE5ObtZx6urqAADOzs6cPrFYjNraWovvQYEjgmfu96dz5swx+6js5o5uAODi4gIAUKvVnD6VSmXsbwoFjgieuStgLZk6Pk/DVFKhUMDb29ukT6FQYNCgQRbfg1YpieBpRNytLQQEBAAACgsLTdrLy8tRVlZm7G8KBY4Ink7E3dpC37590adPHxw+fBg63e8XH7Kzs+Hg4ICIiAiL70FTSiJ4rfVogR07dgAASkpKAAB5eXm4du0aJBIJZs2aBQBYsWIFFi1ahHnz5mHChAn49ddfceDAAchkMvTu3dviMUQGg8FmnvXj5OzLdwk2gX7xbWrbvcNN9m/oOYvTtvL+/hc+zvO+0uXr64v8/Hzj6zNnziA9PR0lJSXo3Lkz3n77bSxevBhOTpbHLxrhiODp0Dpjxu3bt63aLzw8HOHh4c06BgWOCF5zvsrFFwocAJFIhCXJ8UhImIVePbtDoajC0aMnsHrNRtTUWL6Y2V6MXzoV45dOe26/TqNFSt+ZDCuyjlpkM2dFFjUZOLlcjs6dO1t1QU/INqV9hCXJ8cg99jds3vw5Avz7IilpLkJCghARKYMNnea2qYJvr0Bxr4zT3s2/J8IXTkbh2es8VGVZuxnhxo0bh9TUVEyaNIlVPcwFBvZDUuJc5OSeRIxsvrH97r0H2LrlY8hkUTh06BiPFbIjv/UA8lsPOO2ydfXXly4ezuf02YLWOodjocnrcPbwP/t02RQ4ODhg27ZdJu27dh9EdXUNZr7zFk+V2QZnVzFC3xyB/5NXovj8Db7LMUsDA2ezVXZ/4XvI4GDodDpcuWr6YVKpVCgo+DuGDAnhqTLbEDJxGFwlbrh89DwMetv8IOtg4Gy2ymLgRIx+a8QXn25dUFlZZfYLqb/JyyCVvoQOHTrwUJltGB4TBr1ej0tfneO7lOcS0ghncZVy3bp12Lx5s1VvJhKJcObMmRYXxZKbqytUKm7YAKCuTlW/j5srHj/WsCzLJnj38cErQwNw+8dfUFWq4Luc59LacMCeZTFwPj4+6Nq1K4taeFFTWwtvD3ezfS4u4vp97OjSQGPDYsYCsN3Fkga2PIV8lsXAxcXFtetVyofycgQG9IOzszNnWunbrSsUin9Do7G/0c3B0QFD3x6Fp1VK3Dx9he9ymqQx6PkuwWp2v2jy07UCODo6YuhrposjYrEYwcH9ce1aAU+V8SsofDAkUk/8dOxHaNWt9fXgttGuFk3au6+OHIder8eSJfEm7fHzZsDd3Q0HD+XyVBm/hsWEAbD96SQgrMDZ/Ve7CgtvYcen+5CUOBdHvsrEqVP5xm+anD9/AdnZ9hc4ibcXAkaH4N6Nf+Dh7X/xXY5FGghnStlk4G7dusWqDl4tW74a9++XIj5+JiaMH4fKyipkZOzF6jUb7eLi/7P+MHU0HJ0ccfGQ7Y9uAKAT0Dkc/R7OBtHv4UxZ+j3cmz0mctq+eXCyrcppEbufUhLhE9IqJQWOCJ6uvZzDESIEQjqHo8ARwdNS4AhhR0tTSkLYoSklIQxpDcK5yQIFjggejXCEMEQjHCEM0QhHCEMUOEIY0upt+/d6jVHgiODRCEcIQxQ4QhjS6oWzSmn3t1ggwqcz6Dnbi1Kr1di4cSNGjhyJgQMHIiYmBhcvXmz1WilwRPB0ej1ne1ErV65EVlYWJk+ejFWrVsHBwQEJCQn4+eefW7VWChwRPI1ex9lexM2bN3Hy5Em89957WLFiBWQyGbKysuDj44O0tLRWrZUCRwSvpVPKb7/9Fh06dMC0ab8/G08sFmPq1Km4du0aKioqWq1WWjQhgqc3EzClUgmlUslpl0gkkEgkJm3FxcXo3bs33N1N78A9cOBAGAwGFBcXw9vbu1VqtanAadW/8V0CESBVHfdWftu3b0d6ejqnPSkpCcnJySZtCoUCXbp04ewrlUoBgEY4QiyZM2cOoqOjOe3Pjm4AUFdXZ/YJSWJx/bMlVCpVq9VFgSPtkrmp4/O4uLiYfX5EQ9AagtcaaNGE2D2pVGp22qhQ1D+iq7XO3wAKHCHw9/fH3bt3UV1dbdJeUFBg7G8tFDhi9yIjI6HRaHDkyBFjm1qtRk5ODkJDQ80uqDQXncMRuxccHIzIyEikpaVBoVCgR48eyM3NhVwux/r161v1WDb1bAFC+KJSqbBlyxacOHECjx8/hp+fH5YtW4YRI0a06nEocIQwROdwhDBEgSOEIVo0AZCTk4P333/fbN/y5csxf/58xhWx5+fnZ9V+Z8+eRffu3du4mvaLAtdISkoKfHx8TNoCAwN5qoat1NRUk9dZWVmQy+Wc/4g6d+7Msqx2hwLXyOjRoxEQEMB3GbyIiooyeX369Gk8evSI005ahs7hCGGIRrhGlEolqqqqjK9FIhG8vLx4rIi0NxS4RmbPnm3y2s3NrdXvaUHsGwWukTVr1qBHjx7G146OjjxWQ9ojClwjwcHBdrtoQtigRRNCGKLAEcIQBY4QhihwhDBEgSOEIfo9HCEM0QhHCEMUOEIYosARwhAFjhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQxQ4Qhj6fzsSRNhU3tgvAAAAAElFTkSuQmCC\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARy0lEQVR4nO3deVxU5f4H8M+wDZtsCoIogoY4bohLJdc0FQ1RBK4gbhS3pFsKuWXpLTOtXxpiaGIuqEWKyNUkXHIJKarrdsXAuI56IbWIbcgQBYHZfn/wY35MZ2AGhefMge/79Tqvl/Ocw5zv+OLD85xnzswjUqvVahBCmDDhuwBCuhIKHCEMUeAIYYgCRwhDFDhCGKLAEcKQGd8FNCev/JnvEoyCVa9n+C7BqCgafmt1v67fG/Me/TqqnMdiVIEj5JHI6/muwGAUOCJ4aqWC7xIMRoEjwqegHo4QZqiHI4QlChwhDNGkCSEMUQ9HCDtqlZzvEgxGgSPCR0NKQhiiISUhDFHgCGFHTUNKQhiiHo4QhihwhDAkb+C7AoNR4IjwUQ9HCEMK6uEIYUdBPZzRuf1LMY6fzsa5S1fwa0kp6uvl6OPuhikTxyJqVhisrSy1js8rkGL35+mQ3ixE1b37cO7hhKdGDseCqFno4+7G06voeCKRCK/FLUBMzHx49u0NmewuDh8+hjVrN6K29iHf5emmVPJdgcG6TOAyTpxB2hfHMWHsU5g2ZQLMzMxw6cpVbN31OU5nf48DuxJhKRYDAH64cBkLV6xBH3c3zJk5Aw72dii6dQeHj57E19/+gIx929HTuQfPr6hjbEp4F6/FLUDGl18hMXEnJAO9ERv7IoYPH4IpgZEwym/Gp0kT4zP52bFYEBWJbrY2mrbIsGno26cXdqUcxJFjpzE3fAYA4PP0DJiYmGD/jk1wdLDXHN/fqy/e/XALzmR/j6jIMOavoaMNGjQAsYtexJGME5gV+bKm/dbtX7Bl8/uIjAzBwYNf8lhhCwTUw3WZr8kbIhmgFbYmgZPGAQD++/MdTVtNTS3EFuaw62ardaxLDycAgNWfhp+dxezIUJiYmODjj3drte/ecwA1NbWYN+evPFWmh1LB3YyUQYFLTk5GUVFRR9fCi/KKSgBAdycHTZv/UyNRU/sQ/3h/E67/92eUyyrxr4u52Ji0G/08+2DqpPF8lduhRo30hVKpxKV/52m119fXIz//Pxg1ajhPlbVOLZdzNmNl0JBy06ZNcHV1Rf/+/QEAVVVVmDhxInbu3InRo0d3aIEdSalUYsdnaTAzNcW0yRM07TFRs3D3jypknDiDE2e+0bQ/M2Y0Nq59EzY21nyU2+HcevVEZeVdNDRwr4l+KymDv/9omJubQ25sv9AK4QwpH+kaTq1Wo7a2FgoBTcfq8uGWncgvkGLx36Ph1be3pt3ExBQuzt3x9Cg/TBrnD3s7W/z40zUcOHwUK97ZgI8/XANzs853+WttZYX6et0TEHV1jTcIW1tb4d49IwvcY17DXb16FRkZGbh48SJKSkrg4OAAPz8/LFmyBH379tU69sqVK9i4cSOuXbsGW1tbTJ06FcuXL4eVlZVB5+p8vzUG2rrrcxz44hgiQqYi5vlIrX1v/c8m5BdI8eX+HZqZy4Dxf4GHey+8l5CEzK+yED4jkI+yO1Ttw4dw0XGdCwCWlo3/D0b51sBj9ri7d+/GlStXEBgYCB8fH8hkMqSmpiI0NBSHDx/WjOykUimio6PxxBNPYOXKlSgrK8PevXtRXFyMHTt2GHSuLhm4bXv2Y2dKGkKnTcY7K+K09pWWVeDEmW8wd2awJmxNnpv4DN5LSMLlvJ86ZeBKS8oxSDIAFhYWnGGley9XyGS/G99wEoD6MYeU0dHRSEhIgIWFhaYtKCgIwcHBSE5OxoYNGwAAH330ERwcHLBv3z7Y2DT+YerduzfefvttnD9/HmPGjNF7LoNnKUtLS3H9+nVcv34dhYWFAIDi4mJN2583Y7Vtz35s35uKkKkBWLdyCUQikdb+8srfAQBKlYrzs4r/G7ooBTQN3RaXc/NhamqKJ0drT46IxWL4+g5Gbm4+T5XpoVRytzYYMWKEVtgAwNPTE97e3prJwgcPHuDcuXMIDQ3VhA0AQkJCYG1tjZMnTxp0LoN7uMTERCQmJmq1vfPOO5zj1Go1RCIRpFKpoU/NzPa9qdi+NxXBgZPw3j+WwsSE+/fGy6M3TE1NkP3deSz+e7TWWwOZX30NoPEths7on4eOYuWbcXjttQX44V+XNO0LXpoLGxtrHDiYwWN1rWho/7kEtVqNyspKDBw4EABw48YNKBQKDBkyROs4CwsLSCQSg3/fDQrc+vXr21iu8Un74hi27dkPt54ueHrUcJz4+lut/d0dHeD/5AjY23XD/FmhSEk7goi/xWJmcCDs7brhx5+u4cSZb9DH3Q0zg5/j50V0sIKC6/hk+2eIXfQiDv0zGSdPZmvuNMnJOYe0NCMNnI4erbq6GtXV1Zx2Ozs72NnZ6X3Ko0ePory8HEuXLgUAyGQyAICzszPnWGdnZ+Tl5XHadTEocGFhwr+rokB6EwBQWl6Bt97fxNk/ym8o/J8cAQB4fdECeHn0xhfHTiN5XzrkDXK4OHdHZNg0LHxxHmxtdE8sdAbLlq/BnTvFWLBgHoKmTkJl5V1s2/Yp1qzdaJy3dUH3NVxKSgqSkpI47bGxsYiLi+O0N1dUVIR169Zh5MiRCAkJAQDU1dUBAGfoCTQOuZv26yNSG9H/Iq0P14jWh9Omb324B8tmcNpU7+5/pB5OJpNhzpw5UKlUSE9P1/Rop06dwuLFi3Hw4EH4+flp/czixYuRl5eHnJwcva+lS85Sks5FreBOcBk6dGzu/v37iImJwf3795GWlqY1fGz6d9PQsjmZTAYXFxeDztFl7qUknViDgru1UX19PV555RXcvn0bO3fuRL9+2iuoDhgwAGZmZigoKNA+dUMDpFIpJBKJQeehwBHBUytVnK0tlEollixZgry8PGzZsgXDh3PvGe3WrRvGjBmDzMxM1NTUaNozMzNRW1uLwEDD3pelISURPF1DyrbYsGEDsrOzMWHCBFRVVSEzM1Ozz8bGBgEBAQCApUuXYvbs2YiKikJERATKysrw6aefYty4cfD39zfoXDRpYoRo0kSbvkmTqnkTOW0OqdkGP39UVBQuXbqkc5+7uzuys///uS5fvoyEhATNvZRBQUFYtmwZrK0Nu6GdAmeEKHDa9AYucgKnzSH9Gx1H8o+GlETw1Aqj6TP0osARwVM1UOAIYUYtoI9lUuCI4FHgCGFIJRfpP8hIUOCI4KkUFDhCmFEpKXCEMKOkISUh7KgUwrklmAJHBE9JgSOEHSVdwxHCjkpJPRwhzChoSEkIOyoVDSkfSQ/PyXyXYBS8Hdz5LkFQlCrq4QhhRkHXcISwo1TTkJIQZmhISQhDcurhCGFHKaBve6TAEcFTgno4QpiRU+AIYUchosARwoyQ1qOlwBHBkwuohxPO9A4hLVCIRJytrSoqKpCQkICoqCj4+fnBx8cHFy9e1Hns2bNnERYWhqFDh+LZZ59FUlISFArDvjqMAkcETy7ibm1169YtJCcno7y8HD4+Pi0el5OTg0WLFsHe3h6rV69GQEAAtm3bZvCy3DSkJILXHp8/HTx4MC5cuABHR0dkZWVh0aJFOo+Lj4/HoEGDsGfPHpiamgJoXGFn165diIqKgqenZ6vnoR6OCJ5Cx9ZWtra2cHR0bPWYwsJCFBYWIjIyUhM2AJg7dy5UKhXOnDmj9zzUwxHB0zWErK6ufqQ1vltz7do1AMCQIUO02nv27AlXV1fN/tZQ4Ijg6RpSpqSkICkpidMeGxuLuLi4RzpP0/rezdf+buLs7IyKigq9z0GBI4KnawgZ88ILCAsL47Q/au8GAHV1dQAACwsLzj6xWIyHDx/qfQ4KHBE8XUPKxxk6tsTS0hIA0NDQwNlXX1+v2d8amjQhgqeEmrN1hKahZNPQsjmZTAYXFxe9z0GBI4Kn1LF1BIlEAgAoKCjQai8vL0dZWZlmf2u6/JBy2fJX4Dt8MIYPHwJPLw/cuVOMYYPH810WLzz7e2Dh6wswaKgPXFydYWZmhtLfyvDd2XPYm7QPsorf+S5RpwYRmxVQvb290a9fP6SnpyM8PFzz1kBaWhpMTEwwZcoUvc/RauBKSkrg5ORk0NhUqNasXYG7v/+B/Pz/wN6+fcf8QuPaqyecXXog66tvUV5aAYVCiQGSJxARFYqg0MkImzgfdyv/4LtMjvbq0T755BMAQFFREQAgMzMTubm5sLOzw/z58wEAb7zxBl599VW89NJLCAoKws2bN5GamorIyEh4eXnpPYdIrVa3+OdBIpEgPj4ewcHB7fF69LK37c/kPM15evbB7du/AgDOXzoJGxtr3nu4XtbdeT3/nz0XPAmb96xHwrqt2JO0j/n5pRWXWt3/huccTlv87bQ2n6elW7rc3d2RnZ2teZyVlYWkpCQUFRXByckJM2fOxMKFC2Fmpn/A2OoRrWSx02gKG2lZSXEpAMDOvhvPlegmb6dJkhs3bhh0XEBAAAICAh7pHF3+Go5wWYgtYG1jBbFYjP4+Xli+OhYA8F3WOZ4r062jZiU7gt7AiQT0WSPSPsLnhWD1hhWax8V3SrDi1dXIvZjHY1Uta68ejgW9gfvggw+QmJho0JOJRCJkZWU9dlGEX2dP5uBW4W1Y21hDMnQAJjw3Do5ODnyX1SJFZwqcm5sbXF1dWdRCjER5aQXKSxvvCzx7Mgdnjn+DQ6c/g6WVJZI/TuG5Oq5ONaSMjo5mNktJjNPNa4WQ/nQDc/4WbpSBk6tVfJdgMJo0IQYRW1nC3sE436cUUg9Ht3YRjR4uut//e/IvI+E9sB/ycwt07ucbq3sp20OX7+EiZ4eij0fjemw9ejjBwtwcr7/R+PH6X3/5DekHv+SzPKbWxL8JZ5fuuPDDZZQUl0EstsBgXwmmhk5GzYNaxL+7me8SdZJDOEPKVu80YY2PO02On0zFM888rXPf999fwPSp8xhXxN+dJoEzAhAyKwg+g73h1N0BarUaJcVlOJdzCXu37UPpb+W81KXvTpMQj+mctsxfjndUOY+lywfOGBnbrV180xe46R7TOG3HfznRUeU8li4/pCTCR7OUhDCkFNA1HAWOCJ6SejhC2FFQ4AhhR0FDSkLYoSElIQwp1MJZIY4CRwSPejhCGKIejhCGqIcjhCEKHCEMKVSPsiIcPyhwRPCohyOEIQocIQwpVMKZpaSvWCCCp1SrOFtbNTQ0YOPGjRg7diyGDRuGWbNm4fz58+1eKwWOCJ5SpeJsbbVy5UqkpKRgxowZeOutt2BiYoKYmBj8+OOP7VorfeLbCNEnvrXp+8S3rTV31ZoHtbcMfv6rV68iIiICq1atQnR0NIDGFU2nT58OFxcXpKamtqne1lAPRwTvcYeUp06dgrm5OSIiIjRtYrEY4eHhyM3NRUVFRbvVSpMmRPBUOgJWXV2N6upqTruutb+lUim8vLxgY2Oj1T5s2DCo1WpIpVKDlhM2hFEF7t6DIr5LIAJUX8ddcmzr1q1ISkritMfGxiIuLk6rTSaToWfPnpxjm9b0ph6OED1eeOEFhIWFcdr/3LsBQF1dHczNzTntYrEYQOP1XHuhwJFOSdfQsSWWlpaQy+Wc9qagNQWvPdCkCenynJ2ddQ4bZTIZALTb9RtAgSMEAwcOxK1bt1BTU6PVnp+fr9nfXihwpMsLDAyEXC7HoUOHNG0NDQ04cuQIRowYoXNC5VHRNRzp8nx9fREYGIiEhATIZDJ4eHggIyMDJSUlWL9+fbuey6juNCGEL/X19di8eTOOHTuGe/fuwcfHB8uWLYO/v3+7nocCRwhDdA1HCEMUOEIYokkTAEeOHMGqVat07lu+fDlefvllxhWx5+PjY9BxZ8+eRe/evTu4ms6LAtfM0qVL4ebmptU2aNAgnqphKz4+XutxSkoKSkpKOH+InJycWJbV6VDgmhk/fjwkEgnfZfAiJCRE6/Hp06dRVVXFaSePh67hCGGIerhmqqurcffuXc1jkUgER0dHHisinQ0Frpnnn39e67G1tXW7f6cF6doocM2sXbsWHh4emsempqY8VkM6IwpcM76+vl120oSwQZMmhDBEgSOEIQocIQxR4AhhiAJHCEP0eThCGKIejhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQxQ4QhiiwBHCEAWOEIYocIQw9L946k8T2W1ZWAAAAABJRU5ErkJggg==\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQsklEQVR4nO3de1RTV74H8G94v4xAGwSx+GiRlwoqauU6PimDehG4FXCsFKaV1lbwgrReO23ttdOlVlG0glbxxVgL1gfiY6y9iOPSUdErVYYKemV8DMMr1GoUJO/7hzVDPJEECDs54fdZ66zV7H3I+dWVb/Y+Oyc5ArVarQYhhAkrUxdASG9CgSOEIQocIQxR4AhhiAJHCEMUOEIYsjF1Ae3Jm/9u6hLMgmP/35i6BLOikP2zw35drxvbF4f0VDndYlaBI6RL5FJTV2AwChzhPbVSYeoSDEaBI/ynoBGOEGZohCOEJQocIQzRogkhDNEIRwg7apXc1CUYjAJH+I+mlIQwRFNKQhiiwBHCjrqbU8qKigoUFRWhrKwMdXV1cHV1xciRI5Geno6BAwdq7VteXo41a9bg2rVrcHFxwfTp05GZmQlHR0eDjkWBI/zXzRFu27ZtKC8vR2RkJPz8/CAWi7Fnzx7ExMRg//79ePnllwEAVVVVSE5OxiuvvIKlS5eioaEBO3bsQG1tLb7++muDjkWBI/zXzcAlJycjKysLdnZ2mrYZM2YgKioKeXl5WLVqFQBg3bp1cHV1xe7du+Hs7AwAGDBgAD755BOcP38e48eP13ss+j4c4T+5jLt1wqhRo7TCBgCDBg2Cr68vampqAACPHj3CuXPnEBMTowkbAERHR8PJyQnHjx836FgUOMJ/SgV36ya1Wo3m5ma4ubkBAK5fvw6FQoFhw4Zp7WdnZ4eAgABUVVUZ9Lw0pST8p+COaBKJBBKJhNMuFAohFAr1PuXhw4fR2NiIjIwMAIBYLAYAiEQizr4ikQhXrlwxqFQKHOE/BXdEy8/PR05ODqc9NTUVaWlpHT5dTU0NPv/8c4wePRrR0dEAgLa2NgDgTD0BwN7eXtOvT68J3O27tTh6ohTnLpbjH3X1kErleMnbCxFTJyAxPhZOjg7P/dvCoqP4IisXAHDmWCHcXPuyKps5gUCARWnzkZIyD4MGDoBYfA/79x/BZ8vXoLX1sanL002p5DQlJSUhNjaW065vdBOLxXj33XfRt29fbNiwAVZWT866HByevD5kMu5oKpVKNf369JrAFR37AQUHjmLKhHGYGTEFNjY2uFhegY1b/4QTpWfw7dZsONjbc/6uSfwz1m/eCSdHR7Q+NtMXnBGtzfpvLEqbj6JDf0Z29hYE+PsiNfUthIQMQ0RkAszyl/F1LJIYOnVs7+HDh0hJScHDhw9RUFCgNX18+t9Pp5bticVieHh4GHSMXhO41yZPwPzEBPRx+dcKU0LsTAx8qT+25hfi4JETmDt7FufvvliXi5e8vfDy4IE4eqKUZcnMBQYORerCt3Cw6BjiE97RtN+6fRcb1n+BhIRoFBYeMmGFz6FjhOssqVSKBQsW4Pbt29i1axeGDNH+EaKhQ4fCxsYGlZWViIiI0LTLZDJUVVUhKirKoOP0mlXKYQFDtcL2VOS0iQCA//v7HU5fyem/4i9ny7DswzRYW1n+P9WchBhYWVnhq6+2abVv2/4tWlpa8cbv/sNElenRzVVKpVKJ9PR0XLlyBRs2bEBISAhnnz59+mD8+PEoLi5GS0uLpr24uBitra2IjIw06FgGjXB5eXmYOnWq5hN3S9LY1AwAeMHdVav9UUsLVqzbjLjo6Rge6IfCg0dNUR5ToaODoVQqcfGS9oqbVCrF1as/ITSU+0I0B2p5976es2rVKpSWlmLKlCm4f/8+iouLNX3Ozs4IDw8HAGRkZGDOnDlITExEXFwcGhoasHPnTkycOBFhYWEGHcugwK1duxaenp6awN2/fx9Tp07Fli1bMGbMmM7+/5kNpVKJr3cVwMbaGjNfm6LVt27TDqjUKqQv+L2JqmPPq38/NDff07kw8M+6BoSFjYGtrS3k3XyBG52ie1PK6upqAMCpU6dw6tQprT5vb29N4IKCgrBz505kZWVh5cqVcHFxQXx8PBYvXmzwsbp0DqdWq9Ha2gqFjuVYPvlywxZcrazCf76bjMEDB2jayyt+wr7i4/jysyU6p6GWysnREVKp7qs02tqeXCDs5OSIBw/MLHDdPIfbvXu3wfuGhoaisLCwy8fqNYsmz9q49U/49sARxEVPR8qbCZp2uVyO5V9+hVdDQzDjtcmmK9AEWh8/hsdz3mAcHJ6s4JrlRwPmNuJ2oFcGLnf7N9iSX4CYma9h2YfaH4IWHDiKW3dr8WFaCu7W1mnaW359odXWN+BRSyte8vZiWjML9XWNCAwYCjs7O8600ru/J8Tin81vOglA3c0pJUsGB66+vl4z13348CEAoLa2VtP2LH9/fyOUZ3y527/B5h17ED09HJ8vTYdAINDqr2tohEqlwoLMT3X+/e/mp8PR0QGXSopYlMvU/16+ioiIyRg7JgRn/3pR025vb4/g4CCcOXPBhNV1wAgfC7BicOCys7ORnZ2t1bZs2TLOfmq1GgKBwOCLOVnavGMPNu/Yg6jIafjjHzI0VxG0FzMzAqOCgzjtBQeO4tKPFfjjHzIg7OPColzmvtt3GEv/Kw2LFs3XCtz8t+fC2dkJ3xaa6ZuMjD9rCQYFbuXKlT1dR48rOHAEudu/gVc/D7waGoJj//MXrf4X3FwRNnYU/H2HwN+Xe+eV07++ACf/2ziLvbSrsrIamzbvQurCt7DvuzwcP16qudLk9OlzKCgw08BZ2gin65o0vqmsugEAqG9swsdfrOX0h44cjrCxo1iXZXYWZ36GO3dqMX/+G5gxfRqam+8hN3cnPlu+xjwv6wK/zuEEajP6V6T7wz1B94fTpu/+cI8Wcy/Jc1l3uKfK6ZZeuUpJLItaoTJ1CQajwBH+s7RFE0LMmVpJIxwhzNCUkhCG1DIKHCHsKMxmoV0vChzhPTUFjhB2VDIKHCHMqPnzqQAFjvAfBY4QhlRygf6dzAQFjvCeSkGBI4QZlZICRwgzSppSEsKOSsGfH+mlwBHeU1LgCGFHSedwhLCjUtIIRwgzCppSEsKOSkVTyi7xGBShf6deYBH9iFCnKFU0whHCjILO4QhhR6mmKSUhzNCUkhCG5Dwa4fjz1kDIcyhhxdk6q6mpCVlZWUhMTMTIkSPh5+eHsrIynfuePHkSsbGxGD58OCZPnoycnByDb05KgSO8p4SAs3XWrVu3kJeXh8bGRvj5+T13v9OnT2PhwoXo27cvPv30U4SHhyM3N9fgG97QlJLwnrwLAXtWUFAQLly4ADc3N5SUlGDhwoU691u9ejUCAwOxfft2WFtbAwCcnZ2xdetWJCYmYtCgQR0eh0Y4wnsKgYCzdZaLiwvc3Nw63OfmzZu4efMmEhISNGEDgLlz50KlUuGHH37Qexwa4Qjv6bpZlUQigUQi4bQLhUIIhcIuHefatWsAgGHDhmm19+vXD56enpr+jlDgCO/JdYxo+fn5yMnJ4bSnpqYiLS2N024IsVgMABCJRJw+kUiEpqYmvc9BgSO8p2sKmZSUpPNGol0d3QCgra0NAGBnZ8fps7e3x+PHj/U+BwWO8J6uX1joztTxeRwcHAAAMpmM0yeVSjX9HaFFE8J7SgF36wlPp5JPp5bticVieHh46H0OChzhPYWOrScEBAQAACorK7XaGxsb0dDQoOnvCAWO8J5cwN16gq+vL4YMGYK9e/dCqfzX2mhBQQGsrKwQEaH/62V0Dkd4z1hTyE2bNgEAampqAADFxcW4fPkyhEIh5s2bBwBYsmQJ3nvvPbz99tuYMWMGbty4gT179iAhIQGDBw/WewyBWq02m1uPuLm8YuoSzMLvXww1dQlmZd3twg771/rM47Rl3v2m08d53iVd3t7eKC0t1TwuKSlBTk4Oampq4O7ujtdffx3vv/8+bGz0j180whHeM9YU8vr16wbtFx4ejvDw8C4dgwJHeE8Js5mk6UWBI7yn69Iuc9XrA5eRuQAjQoIQEhKEQYN9cPdOLYKDJpu6LJP4bfps/DZ99nP7lXIFPvTlni+ZmkxgISNcXV0d3N3dDfoEna+WLf8A937+BVev/oS+fY17ZQLfVHx/Ec23GzjtXv4+mLpgFn46WW6CqvSzmBFu2rRpWL16NaKioljVw1zIsCm4c/sfAIBzF/8MZ2cnE1dkOvXVd1FffZfTHrdiPgCgbO8p1iUZxGLO4czoE4Me8zRsRDc7R3uE/HsYfqn7GdWnr5i6HJ3kPAocXWlCOhQ881U4Cp1waf9pqFXm+cJWQs3ZzJXeRRNBF749SyzHuPgpUKlUuPideU4nAX6NcHoDt2LFCmRnZxv0ZAKBACUlJd0uipgH0RAvDBnrjxtn/4Z7tdwr5M2FwpIC5+XlBU9PTxa1EDMzLn4KAODC3lI9e5qWOU8hn6U3cMnJyRa9Skl0s7K2QujrE/HongR/O3HJ1OV0SK5WmboEg9GiCdEpKHw0hCJXlB86C6Wsp75hZhwWtWhCeqexv04nzfWzt/bMOWDP6vWBS5gTgwE+/QEAL7zoDjtbW2QueR8AUHu3DnsLD5myPJMQerjBf1Iw7ly5ifrr5v85pRz8mVJ2GLjq6mpWdZjMvKQ4TPjNOK22T5YtBgCcPVPWKwM3ZvYkWNtYo6zQvBdLnlLy6Byu149wUdPfMHUJZufkpkM4uYk/bzQ0pSSEIT6tUlLgCO8pLeUcjhA+oHM4QhhSUOAIYUdBU0pC2KEpJSEMKdT8+ZEFChzhPRrhCGGIRjhCGKIRjhCGKHCEMKRQmff39dqjwBHeoxGOEIYocIQwpFDxZ5WSftOE8J5SreJsnSWTybBmzRpMmDABI0aMQHx8PM6fP2/0WilwhPeUKhVn66ylS5ciPz8fs2bNwscffwwrKyukpKTgxx9/NGqtFDjCe3KVkrN1RkVFBY4dO4YPPvgAS5YsQUJCAvLz8+Hl5YWsrCyj1kqBI7zX3Snl999/D1tbW8TFxWna7O3tMXv2bFy+fBlNTU1Gq5UWTQjvqXQETCKRQCKRcNqFQiGEQu37AFZVVWHw4MFwdnbWah8xYgTUajWqqqrg4eFhlFrNKnC/PLpp6hIID0nbuD/lt3HjRuTk5HDaU1NTkZaWptUmFovRr18/zr4ikQgAaIQjRJ+kpCTExsZy2p8d3QCgra0Ntra2nHZ7e3sAgFQqNVpdFDhikXRNHZ/HwcEBcrmc0/40aE+DZwy0aEJ6PZFIpHPaKBY/uUWXsc7fAAocIfD398etW7fQ0tKi1X716lVNv7FQ4EivFxkZCblcjn379mnaZDIZDh48iFGjRulcUOkqOocjvV5wcDAiIyORlZUFsVgMHx8fFBUVoa6uDitXrjTqsQRqtZo/P8xOSA+RSqVYv349jhw5ggcPHsDPzw+LFy9GWFiYUY9DgSOEITqHI4QhChwhDNGiCYCDBw/io48+0tmXmZmJd955h3FF7Pn5+Rm038mTJzFgwIAersZyUeDaycjIgJeXl1ZbYGCgiapha/Xq1VqP8/PzUVdXx3kjcnd3Z1mWxaHAtTNp0iQEBASYugyTiI6O1np84sQJ3L9/n9NOuofO4QhhiEa4diQSCe7du6d5LBAI4ObmZsKKiKWhwLXz5ptvaj12cnIy+m9akN6NAtfO8uXL4ePjo3lsbW1twmqIJaLAtRMcHNxrF00IG7RoQghDFDhCGKLAEcIQBY4QhihwhDBE34cjhCEa4QhhiAJHCEMUOEIYosARwhAFjhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQ/8Pn2jIm3jXxGQAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAASqUlEQVR4nO3de1xUdf7H8ddwGy6KgIKS5jUFzLvmbq6/8kKEtopsGa2J2qbdQFeldW277M92f9oqaRZmSupSGZWmodtPLaVcb2lh6rKiJmmG3AZNRkWGmWF+f/CThc4Io8CZOfB5Ph7n8XC+5zDn4+PBm+/3nPM95+hsNpsNIYQq3JxdgBAtiQROCBVJ4IRQkQROCBVJ4IRQkQROCBV5OLuAmswl3zu7BJfgc9t/ObsEl2KpOF/nenu/N57tujdVOQ3iUoET4paYTc6uwGESOKF5NqvF2SU4TAIntM8iPZwQqpEeTgg1SeCEUJGcNBFCRdLDCaEeW6W5QT9/7NgxNm/ezMGDB8nPzycgIICBAwcye/ZsunTpUmvbw4cPs2TJEo4fP06rVq0YM2YMSUlJ+Pj4OLQvCZzQvgYOKd9++20OHz5MdHQ0YWFhGAwG1q9fz4QJE9i4cSM9evQAICcnh2nTpnHHHXcwf/58CgsLWbt2LXl5ebz11lsO7UsCJ7SvgUPKadOmkZycjJeXV3Xb2LFjGTduHKmpqbzyyisALF26lICAAN599138/PwA6NSpEy+88AIHDhzg7rvvrndfMpdSaJ/VolxuwqBBg2qFDaBr16707NmT3NxcAK5cucL+/fuZMGFCddgAYmJi8PX1Zdu2bQ7tSwInNM9mNimWBn+nzUZJSQmBgYEAnDx5EovFQp8+fWpt5+XlRUREBDk5OQ59rwwphfbZ6dGMRiNGo1HR7u/vj7+/f71fuWXLFoqKipgzZw4ABoMBgODgYMW2wcHBHDlyxKFSJXBC++wELi0tjZSUFEV7YmIiM2fOrPPrcnNzefnllxk8eDAxMTEAlJeXAyiGngB6vb56fX0kcEL7zBWKpqlTf0dsbKyivb7ezWAw8OSTT9KmTRuWL1+Om1vVUZe3tzcAFRXKfZlMpur19ZHACe2z08M5OnSs6fLly8yYMYPLly+Tnp5ea/h4/d/Xh5Y1GQwGQkJCHNqHnDQR2mepUC43yWQy8dRTT3H27FlWrVpF9+61b2Dt1asXHh4eZGdn12qvqKggJyeHiIgIh/YjgRPaZ7Eol5tgtVqZPXs2R44cYfny5QwYMECxTevWrbn77rvJyMjg6tWr1e0ZGRmUlZURHR3t0L5azJDy7Lk8/rEjk/2HDvNjfgEmk5nbO4YSNWo48Q/H4uvznzH439M/5st9Bzl7Lo9S42Xa+LemW+fbeXTieCLv/ZUT/xdNT6fTMWvmdGbMmEzXLp0wGC6yceNW/rxgCWVl15xdnn1Wa4N+/JVXXiEzM5ORI0dy6dIlMjIyqtf5+fkRGRkJwJw5c3jkkUeIj49n4sSJFBYWsm7dOu655x6GDRvm0L50rvSo86Z8psmylWtJ//gfjBz+C/rdGY6HhweHDh9jR+Y/6XVHN95fvQxvvR6ApBcX4a33oke3zgS08afUeJnPvtjLv46fJHF6PE89NqnJ6gTnPtNk6asLmDVzOps/+V+2b/+CiPCeJCQ8xt69h4iKjsMZvy71PdPk2jvPKdp8pixy+Pvj4+M5dOiQ3XUdO3YkMzOz+vM333xDcnJy9VzKsWPHMnfuXHx9fR3aV4sJXHbOKbrc3pHWrfxqtb++Oo3VaR/wpzlPM+mh8Tf8eYvFysOPzyTvfCEHdmzA3d29yWp1VuB69+7FkcO7+CRjGw/HPVHdnvDMYyx/7a9MnpLABx98onpd9QZu3TxFm89ji5uqnAZpMcdwfSJ6KcIGED36HgC++/6HOn/ew8Od9u3acq28HIulYUMYV/VI3ATc3Nx4/fW3a7W/veZ9rl4t49Hf/sZJldWjgVO71OTQMVxqaiqjRo2qnjXdnBQVlwDQNihAsa7UeBmr1cpPpUY+y9zD3oNZDB3UD71eefGzORgyuD9Wq5VDX9eeNWEymTh69N8MGaI8meAKbOaG3Z6jJocC9+qrr9KhQ4fqwF26dIlRo0axatUq7rrrriYtsClZrVbe+ns6Hu7uPHDfSMX6Bx6ZzqXSqulBHu7u3DfiV7yQlKB2maoJva09JSUX7V7cPZ9fyLBhd+Hp6YnZ1X7BNTTiuKWzlDabjbKyMiw3efrV1fxt+SqOZufw+yen0a1LJ8X61xa+gMlUQXHJBT7L3EO5qYKrZdcIClT2hs2Br48PJpP9a1jl5VUTgn19fSgtdbHANfAspZpazDHcz72x+h3e/3grE2PGMGNKnN1thgzoy69+MZjYB6JY+epf8PP1If7pJEqNl1WuVh1l167dcLjs7V11BtclLw2YzcrFRbXIwK1Y8x6r0tKZ8MB9vPSHuiey1hQzJpKSCz+xc/f+JqzOeQryi2jXLsjuBN2Ot3XAYLjgesNJwGaxKhZX5fCQsqCggBMnTgBVc84A8vLyqtt+Ljw8vBHKa3wr1rzHyrXriRkTycvzZ6PT6Rz+2XJT1bCqufZw32QdJSpqBEPvGsDeff+5LqXX6+nf/0727PnKidXVQUNDSocDt2zZMpYtW1ar7aWXXlJsZ7PZ0Ol0Dt+Qp6aVa9ezcu16xkWP5i9/mlM9E7ymsmvlYLPh61v7oTBWq5UPNv0DgP53uuYfk4b6aMMW5v9xJrNmTa8VuOmPT8LPz5f3P9jsxOrqUKGdcwkOBW7RIsev2ruq9I+3smLNe4S2D+GXQwbw6edf1lrfNjCAYUMHce7H80xLnMd9I4fTrXMn/Fu3prjkAts+/5Iz5/KIGRPJ4AF97O9E47KzT/Dmyr+TmPA7NnyUyrZtmUSE9yQx8Xfs3r2f9HQXDVxz6+Hs3VekNdk5pwAoKCrm+b++qlg/ZGBfhg0dRPuQdoy7fzRZx7LZtXs/ZWXXaNXKj/CePXhy2m95IEp5+aA5mZv0Z374IY/p0x9l7JjRlJRcZMWKdfx5wRKnTOtyhCsfs/1ci5napSXyfrja6pvadWWuckpeq6VbmqqcBmkxdwuI5stmqXR2CQ6TwAnta24nTYRwZTar9HBCqEaGlEKoyFYhgRNCPRaXOdFeLwmc0DybBE4I9VRWSOCEUI1NO1cFJHBC+yRwQqio0uz4LVbOJoETmldpkcAJoZpKqwROCNVYZUgphHoqLdp5NI8ETmieVQInhHqscgwnhHoqrdLDCaEaiwwphVBPZaUMKW/JL/pOcXYJLuGx2xx7m6aoYq2UHk4I1VjkGE4I9Vht2hlSaudPgxA3YK10Uyw3q7i4mOTkZOLj4xk4cCBhYWEcPHjQ7ra7du0iNjaWvn37MmLECFJSUhx+dZsETmie2aZTLDfrzJkzpKamUlRURFhY2A232717NwkJCbRp04YXX3yRyMhIVqxY4fDrAGRIKTTP2gj9xp133slXX31FYGAgO3fuJCHB/ptuFy9eTO/evVmzZg3u7u4A+Pn5sXr1auLj4+natWud+5EeTmieFZ1iuVmtWrUiMDCwzm1Onz7N6dOniYuLqw4bwKRJk6isrOSzzz6rdz/SwwnNM9sJmNFoxGg0Ktr9/f3x9/e/pf0cP34cgD59ar89qX379nTo0KF6fV0kcELzLHZeqpmWlkZKSoqiPTExkZkzHX/rbU0GgwGA4OBgxbrg4GCKi4vr/Q4JnNA8ey+rmjp1qt3XrN1q7wZQXl4OYPeVzHq9nmvX6n//uQROaJ7ZTg/XkKHjjXh7ewNQUVGhWGcymarX10VOmgjNs+h0iqUpXB9KXh9a1mQwGAgJCan3OyRwQvPMOuXSFCIiIgDIzs6u1V5UVERhYWH1+rpI4ITmWXXKpSn07NmT7t278+GHH2Kt8V7x9PR03NzciIqKqvc75BhOaF5jPQf2zTffBCA3NxeAjIwMsrKy8Pf3Z/LkyQDMmzePp59+mscff5yxY8dy6tQp1q9fT1xcHN26dat3Hy71ju9BocOdXYJLGKwPdXYJLiX17IY616fcPlnRlvjjeze9nxtN6erYsSOZmZnVn3fu3ElKSgq5ubkEBQXx4IMP8swzz+DhUX//JT2c0LzGGkKePHnSoe0iIyOJjIy8pX1I4ITmaejVAhI4oX0aeg6sBE5onxWXOQ1RLwmc0Dx7U7tcVYsPXJcetzNj7mNE9O1Fu/bt8PD0oPB8Eft2HeCdN9MpKb7g7BJV59umFWMTYhkYNZTA0CDKr5Rz/tQ5tiz9kO++PuHs8hQqdM2kh8vPzycoKMihOWJa1T40hOCQtnyx7Z8U5RuwWq3cEd6d30wez/0xkTwSOY2fLlxydpmqCerYjj98sAC9rzd7P8qk6Pt8fPx96RTehYAOQc4uz65m08ONHj2axYsXM27cOLXqUd2hvVkc2pulaD/81VEWp/6F8XFjSXvzfSdU5hzTl83Czd2NBdFJlBq08YdGS8dwdU7tcqFr4qoryCsEoHVAaydXop6eQyPoOTSCHasyKDVcwt3DHS9v5a0orsaMTbG4qhZ/DHedl94LXz8fvPRedO/VlVnPPw3Avl0HnFyZevqOHATAhfwSEt/+I31GDMTdw52i7/PZ+vpGDn6yx8kV2qelHq7ewOma6FYHVxM76df8ceHc6s/nz+XzfMICvj14zIlVqatD99sAmLLoKYrPFrAuKQV3Lw+ipo9j+muzcPd0Z/+GL51bpB2u3KP9XL2BW7hwIcuWLXPoy3Q6HTt37mxwUc7wxfY9nDl9Dl8/H8L79OSeqOEEBAU4uyxV6f2qTo6VX71G8m8XYDVXzeE4suNrFu5JIfYPkziwcbfLHWpYmlPgQkND6dChgxq1OFVxgYHigqobC7/cvoddn+7m3W2pePvoWffGzU+E1SJzedWdzF9v2VcdNoAy41WO7vyGYQ+OoH332yjMPe+sEu1qVkPKadOmNeuzlDfyXU4uJ7O/Y+LU2BYTuJ8KLwLYPTtZWlzV5tfGT9WaHGG2VTq7BIfJDah10HvraRPYuM/FcGVnjn4HQGBoW8W6wP+/Bme8oHz0nLNZsSkWV9XiA9c22P7F3CHDBtIjvBv/yvq3yhU5z5EdX3Ptchm/nPBf6H3/M9mhTXAAA6KGUpibj+GHQidWaJ+WAtfiLws897dnCQ5py9f7sijIK8JL70VEvzDujxlN2ZUyli5QPtuwuSozXmXDwneZsuhJntv8P+z76AvcvTwY8WgUHp4epP/3GmeXaJcZ7Qwp6wzciROuN2+use3YvJNfT7yfsQ/eT2DbAGw2KDhfyMfvZvDOynQKzxc5u0RV7UnfyZWLRqKfiiEmKQ5bpY3cw6dI/f1ycrMcu0FTbVYNHcO1+B7u862ZfL41s/4NW5Bvdxzi2x2HnF2Gw1x5CPlzLT5wQvu0dJZSAic0z9pcjuGE0AI5hhNCRRYJnBDqsciQUgj1yJBSCBVZbNp5yIIETmie9HBCqEh6OCFUJD2cECqSwAmhIkuldl7nIYETmic9nBAqksAJoSJLpXbOUrb4RywI7bPaKhXLzaqoqGDJkiUMHz6cfv368fDDD3PgQOM/BFgCJzTPWlmpWG7W/PnzSUtLY/z48Tz//PO4ubkxY8YMvv3220atVQInNM9caVUsN+PYsWN8+umnPPvss8ybN4+4uDjS0tIIDQ0lOTm5UWuVwAnNa+iQcvv27Xh6ejJx4sTqNr1ez0MPPURWVhbFxcWNVqucNBGaV2knYEajEaNR+QxNf39//P1rP2s0JyeHbt264edX+yG3/fr1w2azkZOTQ0hISKPU6lKBO1yw19klCA0ylf+oaHvjjTdISVE+4jAxMZGZM2fWajMYDLRv316xbXBwMID0cELUZ+rUqcTGxiraf967AZSXl+Pp6alo1+v1AJhMpkarSwInmiV7Q8cb8fb2xmw2K9qvB+168BqDnDQRLV5wcLDdYaPBUPU2pcY6fgMJnBCEh4dz5swZrl69Wqv96NGj1esbiwROtHjR0dGYzWY2bNhQ3VZRUcGmTZsYNGiQ3RMqt0qO4USL179/f6Kjo0lOTsZgMNC5c2c2b95Mfn4+ixYtatR96Wyu9v5YIZzAZDLx2muvsXXrVkpLSwkLC2Pu3LkMGzasUfcjgRNCRXIMJ4SKJHBCqEhOmgCbNm3iueees7suKSmJJ554QuWK1BcWFubQdrt27aJTp05NXE3zJYGrYc6cOYSGhtZq6927t5OqUdfixYtrfU5LSyM/P1/xhygoyP470YVjJHA13HvvvURERDi7DKeIiYmp9XnHjh1cunRJ0S4aRo7hhFCR9HA1GI1GLl68WP1Zp9MRGBjoxIpEcyOBq2HKlCm1Pvv6+jb6My1EyyaBq2HBggV07ty5+rO7u7sTqxHNkQSuhv79+7fYkyZCHXLSRAgVSeCEUJEETggVSeCEUJEETggVyf1wQqhIejghVCSBE0JFEjghVCSBE0JFEjghVCSBE0JFEjghVCSBE0JFEjghVCSBE0JF/wdTAs1e6AYGMAAAAABJRU5ErkJggg==\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAASX0lEQVR4nO3de1hUdf4H8PdwG26OYHETxEspNxVUrGRd80Isagis6ZhJsgpqASZa/uxqVk+aoqihpnhjDaFQCC+LFuK6tZoWKP5I0mQxQwSGWAS5zH3/MCenMzKjwHfmMJ/X85znab7nMOeTz7zn+z3fOReBWq1WgxDChIWxCyDEnFDgCGGIAkcIQxQ4QhiiwBHCEAWOEIasjF3AveT1/zF2CSbBru+fjV2CSVHIbnS4XtfnxvrRQd1VTqeYVOAIeShyqbErMBgFjvCeWqkwdgkGo8AR/lNQD0cIM9TDEcISBY4QhmjShBCGqIcjhB21Sm7sEgxGgSP8R0NKQhiiISUhDFHgCGFH3ckh5cWLF5GXl4ezZ8+iuroaTk5OGDFiBJYsWYL+/ftrbVtSUoJ169bh0qVLcHR0xOTJk7Fs2TLY2dkZtC8KHOG/TvZwO3fuRElJCcLDw+Hj4wOJRILMzExERUXhwIEDeOyxxwAA5eXliI2NxeOPP44VK1agpqYGu3fvRlVVFT755BOD9kWBI/zXycDFxsYiJSUFNjY2mrYpU6YgIiIC6enpWLNmDQBgw4YNcHJywr59++Dg4AAA8PLywltvvYUzZ85gzJgxevdF18MR/pPLuMsDGDlypFbYAGDAgAEYPHgwKioqAAC3b9/G6dOnERUVpQkbAERGRsLe3h4FBQUG7YsCR/hPqeAunaRWq1FfXw9nZ2cAwOXLl6FQKDB06FCt7WxsbODn54fy8nKD3peGlIT/FNwerampCU1NTZx2kUgEkUik9y0PHTqE2tpaJCcnAwAkEgkAwMXFhbOti4sLLly4YFCpFDjCfwpuj5aRkYG0tDROe2JiIpKSkjp8u4qKCrz33nsYNWoUIiMjAQDt7e0AwBl6AoBQKNSs18dsAnftehWOHC/C6XMl+KX6JqRSOfp5eiBs4ljEzIyGvZ3tff82O+8IPkjZAgD4+mg2nJ16syqbOYFAgMVJcYiPn4MB/b0gkTTgwIHDWLlqHVpb24xdnm5KJadp7ty5iI6O5rTr690kEgkWLlyI3r17Y9OmTbCwuHPUZWt75/Mhk3F7U6lUqlmvj9kELu/ol8g6eAQTxj6JqWETYGVlhXMlF/Hxjr/jeNHX2L8jFbZCIefv6iS/YuO2PbC3s0Nrm4l+4LrQ+pR3sTgpDnlf/AOpqdvh5zsYiYnzEBQ0FGHhYpjknfF1TJIYOnS8V3NzM+Lj49Hc3IysrCyt4ePd/747tLyXRCKBq6urQfswm8A9M34s4mLE6OX4+wyTOHoq+vfrix0Z2cg9fByzn5vG+bsPNmxBP08PPDawP44cL2JZMnP+/kOQmDAPuXlHMVO8QNNeee06Nm38AGJxJLKzvzBihfeho4d7UFKpFIsWLcK1a9ewd+9eDBqkfROiIUOGwMrKCmVlZQgLC9O0y2QylJeXIyIiwqD9mM0s5VC/IVphuyt80jgAwE//+ZmzrvDUv/HPb87indeSYGnR8/+pZomjYGFhgc2bd2q179y1Hy0trXjh+b8aqTI9OjlLqVQqsWTJEly4cAGbNm1CUFAQZ5tevXphzJgxyM/PR0tLi6Y9Pz8fra2tCA8PN2hfBvVw6enpmDhxouYX956ktq4eAPBIHyet9tstLfhwwzbMiJyMYf4+yM49YozymAoeFQilUolz32nPuEmlUpSW/oDgYO4H0RSo5Z27PGfNmjUoKirChAkT0NjYiPz8fM06BwcHhIaGAgCSk5Mxa9YsxMTEYMaMGaipqcGePXswbtw4hISEGLQvgwK3fv16uLu7awLX2NiIiRMnYvv27Rg9evSD/v+ZDKVSiU/2ZsHK0hJTn5mgtW7D1t1QqVVYsuhvRqqOPY++bqivb9A5MXCjugYhIaNhbW0NeSc/4F1O0bkh5Y8//ggAOHnyJE6ePKm1ztPTUxO4gIAA7NmzBykpKVi9ejUcHR0xc+ZMLF261OB9PdQxnFqtRmtrKxQ6pmP55KNN21FaVo5XFsZiYH8vTXvJxR+Qk1+Aj1Yu1zkM7ans7ewgleo+S6O9/c4Jwvb2drh1y8QC18ljuH379hm8bXBwMLKzsx96X2YzafJHH+/4O/YfPIwZkZMR/6JY0y6Xy7Hqo814KjgIU54Zb7wCjaC1rQ2u9/mCsbW9M4Nrkj8NmFqP2wGzDNyWXZ9ie0YWoqY+g3de0/4RNOvgEVRer8JrSfG4XlWtaW/57YNWdbMGt1ta0c/Tg2nNLNysroW/3xDY2NhwhpWefd0hkfxqesNJAOpODilZMjhwN2/e1Ix1m5ubAQBVVVWatj/y9fXtgvK63pZdn2Lb7kxETg7FeyuWQCAQaK2vrqmFSqXComVv6/z75+OWwM7OFt8V5rEol6nvi0sRFjYeT4wOwjf/PqdpFwqFCAwMwNdff2vE6jrQBT8LsGJw4FJTU5GamqrV9s4773C2U6vVEAgEBp/MydK23ZnYtjsTEeGT8P4byZqzCO4VNTUMIwMDOO1ZB4/gu/MX8f4byRD1cmRRLnOf5xzCiv9LwuLFcVqBi5s/Gw4O9tifbaJfMjL+zCUYFLjVq1d3dx3dLuvgYWzZ9Sk83FzxVHAQjn71T631jzg7IeSJkfAdPAi+g7lPXjn12wdw/J+e7LGndpWV/Yit2/YiMWEecj5PR0FBkeZMk1OnTiMry0QD19N6OF3npPFNWfkVAMDN2jq8+cF6zvrgEcMQ8sRI1mWZnKXLVuLnn6sQF/cCpkyehPr6BmzZsgcrV60zzdO6wK9jOIHahP4V6flwd9Dz4bTpez7c7aXcU/IcNxzqrnI6xSxnKUnPolaojF2CwShwhP962qQJIaZMraQejhBmaEhJCENqGQWOEHYUJjPRrhcFjvCemgJHCDsqGQWOEGbU/PlVgAJH+I8CRwhDKrlA/0YmggJHeE+loMARwoxKSYEjhBklDSkJYUel4M9NeilwhPeUFDhC2FHSMRwh7KiU1MMRwoyChpSEsKNS0ZDyodDNc+5Y5THe2CXwilJFPRwhzCjoGI4QdpRqGlISwgwNKQlhSM6jHo4/Xw2E3IcSFpzlQdXV1SElJQUxMTEYMWIEfHx8cPbsWZ3bnjhxAtHR0Rg2bBjGjx+PtLQ0gx9OSoEjvKeEgLM8qMrKSqSnp6O2thY+Pj733e7UqVNISEhA79698fbbbyM0NBRbtmwx+IE3NKQkvCd/iID9UUBAAL799ls4OzujsLAQCQkJOrdbu3Yt/P39sWvXLlhaWgIAHBwcsGPHDsTExGDAgAEd7od6OMJ7CoGAszwoR0dHODs7d7jN1atXcfXqVYjFYk3YAGD27NlQqVT48ssv9e6HejjCe7oeVtXU1ISmpiZOu0gkgkgkeqj9XLp0CQAwdOhQrXY3Nze4u7tr1neEAkd4T66jR8vIyEBaWhqnPTExEUlJSZx2Q0gkEgCAi4sLZ52Liwvq6ur0vgcFjvCeriHk3LlzdT5I9GF7NwBob28HANjY2HDWCYVCtLW16X0PChzhPV13WOjM0PF+bG1tAQAymYyzTiqVatZ3hCZNCO8pBdylO9wdSt4dWt5LIpHA1dVV73tQ4AjvKXQs3cHPzw8AUFZWptVeW1uLmpoazfqOUOAI78kF3KU7DB48GIMGDcJnn30GpfL3udGsrCxYWFggLCxM73vQMRzhva4aQm7duhUAUFFRAQDIz89HcXExRCIR5syZAwBYvnw5XnrpJcyfPx9TpkzBlStXkJmZCbFYjIEDB+rdh0CtVpvMo0esbDyNXYJJoAtQtb35c2aH69d7z+G0Lbv+6QPv536ndHl6eqKoqEjzurCwEGlpaaioqECfPn0wffp0vPzyy7Cy0t9/UQ9HeK+rhpCXL182aLvQ0FCEhoY+1D4ocIT3lDCZQZpeFDjCe7pO7TJVFDgAAoEAi5PiEB8/BwP6e0EiacCBA4exctU6tLbqP3ugJ7G2F2L03/6CgGkh6O31KJQyBRoqb+L8/pO4eOBfxi5PJ5mgh/Rw1dXV6NOnj0G/oPPZ+pR3sTgpDnlf/AOpqdvh5zsYiYnzEBQ0FGHhYpjQvFL3EggwK2M5vEYNwf8f/Be+33scVnZCBEwbg4j1C/HI431xck22savk6DE93KRJk7B27VpERESwqoc5f/8hSEyYh9y8o5gpXqBpr7x2HZs2fgCxOBLZ2V8YsUJ2PEc8Bu8nfHF2ZwEK3/99lq9431dYVJSCkbMnmmjg+POF2OEP3+bwzT5LHAULCwts3rxTq33nrv1oaWnFC8//1UiVsSd0tAMA3K77r1a7Sq5EW0Mz5G1SY5SllxxqzmKqzP4YLnhUIJRKJc59d0GrXSqVorT0BwQHBxmpMvaqL1Sg7VYLnlr4LBp/kaD6QgWs7WwwbPo4uA8biII3dhu7RJ341MPpDZzgIa6e5ROPvm6or2/QeQb4jeoahISMhrW1NeRyuRGqY6u9qRU589dj6kdxmL7tFU27tLkNBxdtxJUvi41Y3f2Zco/2R3oD9+GHHyI1NdWgNxMIBCgsLOx0USzZ29lBKuWGDQDa2+8Moezt7XDrVs8PHADIWtshuVKFK4UluFH8E2ydHBD84jOI2pyAnLgNqPymTP+bMKboSYHz8PCAu7s7i1qMorWtDa6ODjrX2doK72xjJj8NuPj0w9zcd1H43qcoyTyhaf8h/wwWfPURpqyJw9ZxyVCrTOsD3qOGlLGxsT16lvJmdS38/YbAxsaGM6z07OsOieRXsxhOAsCTceGwtrVB+VHt+zEq2mW4WnQeo2P/gt5eLmi8rv9WAizJ1Spjl2Aws7885/viUlhaWuKJ0dqTI0KhEIGBASguLjVSZew5uvUBAAgsuR8Li9/uUmVhZclZZ2xKqDmLqTL7wH2ecwgqlQqLF8dptcfNnw0HB3vsz84zUmXs1f90AwAw/LlxWu1CkT2GhI1CW+Nt/PdajTFK6xCfAmf2PwuUlf2Irdv2IjFhHnI+T0dBQZHmTJNTp04jK8t8Andu9zEMmz4WE1eI4erbD1XfX4GtkwNGPD8BvdycceytPSZ3/AYAcvBnSEnXwwGwsLDAK4vjERf3Agb090J9fQNycu6cS9nS0sq8HmNeD+fk7Yo/vxKNAX8aCodHRVC0y1F76Wec212Ay8e+N0pN+q6Hi/R+ltOWf/1Id5XTKRQ4E0QXoGrTF7hnvady2o5cP9pd5XSK2Q8pCf/xaZaSAkd4T8mjYzgKHOE9JfVwhLCjoMARwo6ChpSEsENDSkIYUqj5c5MFChzhPerhCGGIejhCGKIejhCGKHCEMKRQddcT4boeBY7wHvVwhDBEgSOEIYWKP7OUZn+LBcJ/SrWKszwomUyGdevWYezYsRg+fDhmzpyJM2fOdHmtFDjCe0qVirM8qBUrViAjIwPTpk3Dm2++CQsLC8THx+P8+fNdWisFjvCeXKXkLA/i4sWLOHr0KF599VUsX74cYrEYGRkZ8PDwQEpKSpfWSoEjvNfZIeWxY8dgbW2NGTNmaNqEQiGee+45FBcXo66u6+7DSZMmhPdUOgLW1NSEpqYmTrtIJIJIJNJqKy8vx8CBA+HgoH0H7uHDh0OtVqO8vByurq5dUqtJBU4hu2HsEggPSdt/4bR9/PHHSEtL47QnJiYiKSlJq00ikcDNzY2zrYuLCwBQD0eIPnPnzkV0dDSn/Y+9GwC0t7fD2tqa0y4U3nm2hFTadc/Fo8CRHknX0PF+bG1tdT4/4m7Q7gavK9CkCTF7Li4uOoeNEokEALrs+A2gwBECX19fVFZWoqWlRau9tLRUs76rUOCI2QsPD4dcLkdOTo6mTSaTITc3FyNHjtQ5ofKw6BiOmL3AwECEh4cjJSUFEokE3t7eyMvLQ3V1NVavXt2l+zKpZwsQYixSqRQbN27E4cOHcevWLfj4+GDp0qUICQnp0v1Q4AhhiI7hCGGIAkcIQzRpAiA3Nxevv/66znXLli3DggULGFfEno+Pj0HbnThxAl5eXt1cTc9FgbtHcnIyPDw8tNr8/f2NVA1ba9eu1XqdkZGB6upqzhdRnz59WJbV41Dg7vH000/Dz8/P2GUYRWRkpNbr48ePo7GxkdNOOoeO4QhhiHq4ezQ1NaGhoUHzWiAQwNnZ2YgVkZ6GAnePF198Ueu1vb19l9/Tgpg3Ctw9Vq1aBW9vb81rS0tLI1ZDeiIK3D0CAwPNdtKEsEGTJoQwRIEjhCEKHCEMUeAIYYgCRwhDdD0cIQxRD0cIQxQ4QhiiwBHCEAWOEIYocIQwRIEjhCEKHCEMUeAIYYgCRwhDFDhCGPofraqPafoBehYAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACdCAYAAADfXhZIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQUElEQVR4nO3df3CM974H8PeGZCNxt7Op/OT60V7yq0TU0WlHq+dQcpyiLhUNQgnTysbvotNRNXOGOcT1K5x7pMreuRrEj6JKHcwY7SklU9GQ0KjjV0Q2ctkS2d/3D5Mc69lkl2y+z/Nk36+Z54/9fp8++zGTd7/f57u730fjcrlcICIhguQugCiQMHBEAjFwRAIxcEQCMXBEAjFwRAIxcEQCMXBEAKqqqpCbm4sJEyYgNTUV8fHxOHXqlMdzjx49ipEjR6Jnz5548803kZeXB7vd7tP7MHBEAK5cuYL8/Hzcvn0b8fHxjZ53/PhxZGdn47nnnsOiRYswaNAgrF+/HsuWLfPpfdr6q2AiNUtOTsbJkyeh1+tx5MgRZGdnezxv+fLlSEpKwqZNm9CmTRsAQHh4ODZu3IgJEyaga9euTb4PRzgiAO3bt4der2/ynPLycpSXlyM9Pb0hbACQkZEBp9OJw4cPe30fBo7IRxcuXAAAvPTSS27t0dHRiImJaehvCqeU1CqZzWaYzWZJu06ng06ne6ZrmkwmAEBkZKSkLzIyElVVVV6voajA2ap/lbsERWgX97rcJSiK3XqzyX5PfzfGggPIy8uTtBsMBuTk5DxTHXV1dQCAkJAQSZ9Wq8XDhw+9XkNRgSN6JjaLpGnixIkYOXKkpP1ZRzcACA0NBQBYrVZJn8ViaehvCgNHqudySD8Da87UsTH1U0mTyYSoqCi3PpPJhNTUVK/X4KIJqZ/dIj1aQGJiIgCgpKTErf327duorKxs6G8KA0eq53LYJUdL6N69O1544QVs374dDoejob2goABBQUEYPHiw12twSknq56eAbdiwAQBw+fJlAMDevXtRVFQEnU6H8ePHAwDmz5+PDz/8EFOmTMHQoUNx6dIlbN26Fenp6ejWrZvX99AoaU8TrlI+wlVKd95WKS3nj0ratMkDn/p9GvtKV8eOHXHs2LGG10eOHEFeXh4uX76MiIgIjBo1CtOnT0fbtt7HLwZOgRg4d14Dd+5bSZu215CWKqdZOKUk1XM5bXKX4DMGjtTPw+dwSsXAkfq10KpkS2DgSP0YOCJxXJxSEgnEEY5IIAaOSCCb9Nv7SsXAkfpxhCMSyM4RjkgcH/eEVIKAClz+/2xH6aVyXLhYjhsVlYiLicLhXcZGzz93vgxrNxpx7vxFaDRA755JmP3B+0jo8aLAqsXSaDSYkZOFqVPHo2uXTjCZarBz534sXrICtbXetxCQxWM/lVG6gPo93Jq/bcGpomJ0iouF7t/aN3lucUkpJhnm40ZFJQxZE5A9ZQKuXr+JzOkf4dLlK4IqFm9l7mdYmfsZSksvYeasRdi162sYDJOxd48RGo1G7vI8s1mlh0IF1Ah3cMcX+PeOsQCAd8Z/gNomNn1Ztvq/Edy2LYwbViA6sgMAYMjA1zE8YxpWrMtH/uqlQmoWKSmpBwzZk7F7zwGMSZ/W0H7ln9ewZvWfkZ4+Atu2fSVjhY3gCKdM9WHz5tqNCpSUXsLgP7zeEDYAiI7sgMF/eB0nz5xF9Z2alipTNmPT30FQUBDWrv3crf3zTV/iwYNajHvvP2WqzAuHXXoolE+By8/Pb/gVbCAoKb0EAEhJlu5RkZKcAJfLhfMXy0WX1eL6vpwCh8OBH0+fdWu3WCwoLj6Pvn17y1RZ01w2m+RQKp8Ct3LlSrddZe/evYs+ffrg9OnTLVaYnKqq7wAAoiOfl/RFdXjUVmWqFlqTCLFx0aiurvG4DdzNikpERj6P4OBgGSrzwu6QHgr1TFNKl8uF2tpanx/RozZ1dY++DBsSIv3j0mpD3M5pTcLatYPF4nnBof7fGxbWTmRJvnE4pIdCBdSiia9CQ7UAAKtVOjWp/4OsP6c1qX34EFHtwz321f97FfnRgIKnkE8KqEUTX9VPG2+b7kj66qebUY8tprQWtypuo0OHCI9beXeMi4HJdAc2Bf5xu+wOyaFUPo9wt27dQllZGQDgt99+AwDcuHGjoe1JCQkJfihPHi8l9gAAFJ8vxejhaW59xefLoNFokBz/H3KU1qLOFBVj8OA30e93vfHd9z82tGu1WqSkJOPEiZMyVtcEBU8hn+Rz4FatWoVVq1a5tX366aeS81wuFzQaDUpLS5tfnUw6d4pDckJ3HD52AjlZmYiKrF8ouYPDx07glZdT0OH5CJmr9L8dhfuwcEEOZszIcgtc1pQMhIeH4ctte2SsrglW9awl+BQ4Xx+nqnT7Dh3FrcpHjxSquXsPdrsdf9tSAACIjYnC8LR/7WW4cNYHmJyzAJnT52Hc6OEAgK0798HpcmGeYar44gUoKSnDhr9ugSF7Mgp35OPgwWNITOgOg2Eyjh//BwoKFBo4FY1wAbUv5STDfJz56WePfX1Te2JL3nK3trMlpVi30YhzFy5CAw1690zErA/eR1ILTyfl3JcyKCgIM2dMRVbWOHTt0gnV1TUoLHz0XcoHD2plqcnbvpT3F0g/kG//l90tVU6zBFTg1IIbwbrzGrg5wyVt7f9rX0uV0yz8WIBUz2V3yl2Czxg4Ur/WtmhCpGQuB0c4ImE4pSQSyGVl4IjEsStmod0rBo5Uz8XAEYnjtDJwRMK41POpAANH6sfAEQnktCl0+z4PGDhSPaedgSMSxulg4IiEcXBKSSSO066erXkYOFI9BwNHJI6jmfdwp06dQmZmpse+b775Bi++6L+nJTFwpHpOh39GuIkTJyI5OdmtLTo62i/XrsfAkerZ/TSl7NevHwYNGuSXazVGPZNfokY4nRrJ8azu37/folv4K2qE4+Y5j/wxJlXuElTF4ZSOG2azGWazWdKu0+mg0+k8Xuejjz5CbW0t2rZti1deeQULFixAfHy8X2tVVOCInoXdwz2c0WhEXl6epN1gMCAnJ8etLTg4GEOGDMEbb7wBvV6Pixcv4osvvkBGRgZ27tyJbt26+a1WRW2T1zako9wlKAJHOHf7r33dZP/JOOm+lEllW556hHtcWVkZRo0ahbS0NKxcudL3Yr3gCEeq52lK6WuwGpOQkIBXX30VJ0/693kKXDQh1bO5NJLDH2JjY3Hv3j2/XKseRzhSPUcLjRvXr1+HXq/36zU5wpHqOaCRHE+jpqZG0nbmzBmcOnUK/fv391eZADjCUStge8qAPWnWrFlo164dUlNTodfr8csvv2D79u3Q6/WSFc3mYuBI9eya5gVu0KBB2L9/PzZv3oz79+8jIiICb7/9NnJychAXF+enKh9h4Ej1mvt0uMzMzEa/vOxvDBypnq2ZI5xIDBypXnOnlCIxcKR6KtphgYEj9VPRHkIMHKmfivaBZeBI/TilJBKIU0oigTilJBKIU0oigRxQzG+ovWLgSPWa+9UukRg4ABqNBjNysjB16nh07dIJJlMNdu7cj8VLVqC29qHc5QkT1y0Ovx/5e/R+IxWxXWIRrA1G5dVKfH/gO+zdtBeWhxa5S/TIqlHPCNfk7+EqKipQV1cnqhbZrMz9DCtzP0Np6SXMnLUIu3Z9DYNhMvbuMUKjoq8NNddb6W9heNYIVF6txLY1Bdi8dDNu/noDE+ZnYsWeXIRoQ+Qu0SOHh0OpmhzhBg4ciOXLl2PYsGGi6hEuKakHDNmTsXvPAYxJn9bQfuWf17Bm9Z+Rnj4C27Z9JWOF4nz/zfcoXF+I2t9qG9oO/e9BVFypQPqMsXhr7GAcMDa9oY8c1HQP1+QIp6ANvVrM2PR3EBQUhLVrP3dr/3zTl3jwoBbj3pPuCNValZ8rdwtbvRP7TwAAusR3EV2ST2xwSQ6lCvh7uL4vp8DhcODH02fd2i0WC4qLz6Nv394yVaYcHWI7AADuVv+fzJV41mpGOACt/h4mNi4a1dU1sFqtkr6bFZWIjHwewcHBMlSmDEFBQUifMRZ2mx3HvzoudzketaoRbunSpVi1apVPF9NoNDhy5EizixIprF07WCzSsAFAXd2jVbmwsHa4d88msizFyFo8FYl9E2H8ixE3f70pdzke2RUcsCd5DVxsbCxiYmJE1CKL2ocPEdU+3GNfaKj20TkB9NHA48bNHY9h7w/Doa0HsXN9odzlNEpNU0qvgZs0aVKrXqW8VXEbSYk9EBISIplWdoyLgcl0BzZb4I1u783OwNiZY/H37X/H+o/Xy11Ok2wup9wl+Czg96U8U1SMNm3aoN/v3BdHtFotUlKSUVRULFNl8nlvdgYyZmfgaOERrJu/Vu5yvHLAJTmUKuADt6NwH5xOJ2bMyHJrz5qSgfDwMHy5bY9Mlclj7MyxyJidgWO7jmHNvDWq+GhITYEL+I8FSkrKsOGvW2DInozCHfk4ePAYEhO6w2CYjOPH/4GCgsAJ3NDMP2Hc3PGoulGFs9+dxYB3Brj1362+i7MnzjbyX8vHBvVMKZsMXFlZmag6ZDVn7mJcvXoDWVnjMPSPA1FdXYP16zdj8ZIVqvg/vL/0SOkOAIjqFIU5q+ZI+n/+4WdFBs6hons4Ph9Ogfh8OHfeng/3duc/Sdq+vnagpcpploCfUpL6qWmVkoEj1XO0lns4IjVQ0z0cA0eqZ2fgiMSxc0pJJA6nlEQC2V1K3lTBHQNHqscRjkggjnBEAnGEIxKIgSMSyO5Uz+M8GDhSPY5wRAIxcEQC2Z3qWaUM+C0WSP0cLqfkeFpWqxUrVqxA//790atXL4wZMwY//PCD32tl4Ej1HE6n5HhaCxcuhNFoxPDhw/HJJ58gKCgIU6dOxU8//eTXWhk4Uj2b0yE5nsa5c+dw4MABzJs3D/Pnz0d6ejqMRiNiY2ORm5vr11oZOFK95k4pDx06hODgYLz77rsNbVqtFqNHj0ZRURGqqqr8VisXTUj1nB4CZjabYTabJe06nQ46nc6trbS0FN26dUN4uPsO3L169YLL5UJpaSmioqL8UquiAme3KnPvelI2S911Sdu6deuQl5cnaTcYDMjJyXFrM5lMiI6OlpwbGRkJABzhiLyZOHEiRo4cKWl/cnQDgLq6Oo9PSNJqHz1bwmLx36OWGThqlTxNHRsTGhrq8fkR9UGrD54/cNGEAl5kZKTHaaPJZAIAv92/AQwcERISEnDlyhU8ePDArb24uLih318YOAp4aWlpsNlsKCz81zPwrFYrdu/ejT59+nhcUHlWvIejgJeSkoK0tDTk5ubCZDKhc+fO2LNnDyoqKrBs2TK/vpeini1AJBeLxYLVq1dj//79uHfvHuLj4zFnzhy89tprfn0fBo5IIN7DEQnEwBEJxEUTALt378bHH3/ssW/u3LmYNm2a4IrEi4+P9+m8o0ePolOnTi1cTevFwD1m9uzZiI2NdWtLSkqSqRqxli9f7vbaaDSioqJC8j+iiIgIkWW1OgzcYwYMGIDExES5y5DFiBEj3F5/++23uHv3rqSdmof3cEQCcYR7jNlsRk1NTcNrjUYDvV4vY0XU2jBwj8nMzHR7HRYW5vc9LSiwMXCPWbJkCTp37tzwuk2bNjJWQ60RA/eYlJSUgF00ITG4aEIkEANHJBADRyQQA0ckEANHJBB/D0ckEEc4IoEYOCKBGDgigRg4IoEYOCKBGDgigRg4IoEYOCKBGDgigRg4IoH+H9ffABBhtA5sAAAAAElFTkSuQmCC\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "fy9bh1zk2vNH",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "54c8b17f-69c2-40c9-df1d-93ca48fecbfc"
+ },
+ "source": [
+ "# Thus in binary classification, the count of true negatives is C0.0, false negatives is C1.0, true positives is C1.1 and false positives is C0.1.\n",
+ "len(confusion_matrix_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "76"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 63
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "6XEYkdWt7qdT"
+ },
+ "source": [
+ "# importing mean() for calculate the mean of all the lists (precision, recall, f1-score ... ) \n",
+ "from statistics import mean \n",
+ " "
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "L3XL1UL9tAF-",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "16a95516-2558-4106-9a11-ec582004c774"
+ },
+ "source": [
+ "mean(roc_auc_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.943560240928662"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 65
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "vmwc4-L5rCQr",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "65c16acc-069b-4258-e323-2d0745a27ad3"
+ },
+ "source": [
+ "mean(recall_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.887120481857324"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 66
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "LodlrZmR-J9e",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "7e8ec3e6-38e9-458a-a114-70dfc73d985b"
+ },
+ "source": [
+ "mean(accuracy_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9695723684210527"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 67
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "rPa9hk3joMmZ",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "39fdf5f7-0a0b-4661-8c40-083bb9332705"
+ },
+ "source": [
+ "mean(precision_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "1.0"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 68
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "fW1Lwwi3nW4U",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "fa5efa5f-456c-4d22-f8cb-fac08fc96784"
+ },
+ "source": [
+ "mean(f1_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9363818332706881"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 69
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "xk3w__yQW01i",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "efc03b83-1458-4df2-973b-d0edab85a86e"
+ },
+ "source": [
+ "mean(matthews_set)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9219909077922733"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 70
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "_pe4J3UxW7Ha",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "2a37f9d3-437e-454a-955d-7a66fe9bc1e1"
+ },
+ "source": [
+ "# Combine the predictions for each batch into a single list of 0s and 1s.\n",
+ "flat_predictions = [item for sublist in predictions_test for item in sublist]\n",
+ "flat_predictions = np.argmax(flat_predictions, axis=1).flatten()\n",
+ "\n",
+ "# Combine the correct labels for each batch into a single list.\n",
+ "flat_true_labels = [item for sublist in true_labels for item in sublist]\n",
+ "\n",
+ "# Calculate the MCC\n",
+ "mcc = matthews_corrcoef(flat_true_labels, flat_predictions)\n",
+ "\n",
+ "print('MCC: %.3f' % mcc)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "MCC: 0.919\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "vz6AKXVLw-Kz"
+ },
+ "source": [
+ "#6. Test on an entire novel sentences (the novel \"maison\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "-zVTp2oNxzC6"
+ },
+ "source": [
+ "##Download the novels"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "eHNYGQdBOsIR"
+ },
+ "source": [
+ "We must create a new folder in (content) on colab and we name it (Romans). Then we upload the novels we want"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "K8Xs9F4N43SZ"
+ },
+ "source": [
+ "from os import listdir\n",
+ "from os.path import isfile, join\n",
+ "dir = \"/content/Romans\"\n",
+ "onlyfiles = [f for f in listdir(dir) if isfile(join(dir, f))] # list of novels names(with extentions filename.txt) in the directory path"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "ID-xnWgo5Cyx",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "5af26370-bf86-4fa6-f2b1-e15d0cc5ce9b"
+ },
+ "source": [
+ "onlyfiles[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "'maison.txt'"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 51
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "KS3zVoeHvQJR"
+ },
+ "source": [
+ "# Put the novels contents in list\n",
+ "content_french = [] #List of lists of sentences of novels\n",
+ "file_content =[]\n",
+ "for file in onlyfiles:\n",
+ " #file_content.append( file.read )\n",
+ " f = open('/content/Romans/'+ file)\n",
+ " #file_content.append(f)\n",
+ " content_french.append(f.read())"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "km8oomkc75kA",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 55
+ },
+ "outputId": "24ea0245-f063-4682-e479-d2539b6d95b7"
+ },
+ "source": [
+ "content_french[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "'\\n\\n\\nAu milieu de la rue Saint-Denis, presque au coin de la rue du Petit-Lion, existait naguère une de ces maisons précieuses qui donnent aux historiens la facilité de reconstruire par analogie l’ancien Paris. Les murs menaçants de cette bicoque semblaient avoir été bariolés d’hiéroglyphes. Quel autre nom le flâneur pouvait-il donner aux X et aux V que traçaient sur la façade les pièces de bois transversales ou diagonales dessinées dans le badigeon par de petites lézardes parallèles ? Evidemment, au passage de toutes les voitures, chacune de ces solives s’agitait dans sa mortaise. Ce vénérable édifice était surmonté d’un toit triangulaire dont aucun modèle ne se verra bientôt plus à Paris. Cette couverture, tordue par les intempéries du climat parisien, s’avançait de trois pieds sur la rue, autant pour garantir des eaux pluviales le seuil de la porte, que pour abriter le mur d’un grenier et sa lucarne sans appui. Ce dernier étage était construit en planches clouées l’une sur l’autre comme des ardoises, afin sans doute de ne pas charger cette frêle maison.\\n\\nPar une matinée pluvieuse, au mois de mars, un jeune homme, soigneusement enveloppé dans son manteau, se tenait sous l’auvent de la boutique qui se trouvait en face de ce vieux logis, et paraissait l’examiner avec un enthousiasme d’archéologue. A la vérité, ce débris de la bourgeoisie du seizième siècle pouvait offrir à l’observateur plus d’un problème à résoudre. Chaque étage avait sa singularité. Au premier, quatre fenêtres longues, étroites, rapprochées l’une de l’autre, avaient des carreaux de bois dans leur partie inférieure, afin de produire ce jour douteux, à la faveur duquel un habile marchand prête aux étoffes la couleur souhaitée par ses chalands. Le jeune homme semblait plein de dédain pour cette partie essentielle de la maison, ses yeux ne s’y étaient pas encore arrêtés. Les fenêtres du second étage, dont les jalousies relevées laissaient voir, au travers de grands carreaux en verre de Bohême, de petits rideaux de mousseline rousse, ne l’intéressaient pas davantage. Son attention se portait particulièrement au troisième, sur d’humbles croisées dont le bois travaillé grossièrement aurait mérité d’être placé au Conservatoire des arts et métiers pour y indiquer les premiers efforts de la menuiserie française. Ces croisées avaient de petites vitres d’une couleur si verte, que, sans son excellente vue, le jeune homme n’aurait pu apercevoir les rideaux de toile à carreaux bleus qui cachaient les mystères de cet appartement aux yeux des profanes. Parfois, cet observateur, ennuyé de sa contemplation sans résultat, ou du silence dans lequel la maison était ensevelie, ainsi que tout le quartier, abaissait ses regards vers les régions inférieures. Un sourire involontaire se dessinait alors sur ses lèvres, quand il revoyait la boutique où se rencontraient en effet des choses assez risibles. Une formidable pièce de bois, horizontalement appuyée sur quatre piliers qui paraissaient courbés par le poids de cette maison décrépite, avait été rechampie d’autant de couches de diverses peintures que la joue d’une vieille duchesse en a reçu de rouge. Au milieu de cette large poutre mignardement sculptée se trouvait un antique tableau représentant un chat qui pelotait. Cette toile causait la gaieté du jeune homme. Mais il faut dire que le plus spirituel des peintres modernes n’inventerait pas de charge si comique. L’animal tenait dans une de ses pattes de devant une raquette aussi grande que lui, et se dressait sur ses pattes de derrière pour mirer une énorme balle que lui renvoyait un gentilhomme en habit brodé. Dessin, couleurs, accessoires, tout était traité de manière à faire croire que l’artiste avait voulu se moquer du marchand et des passants. En altérant cette peinture naïve, le temps l’avait rendue encore plus grotesque par quelques incertitudes qui devaient inquiéter de consciencieux flâneurs. Ainsi la queue mouchetée du chat était découpée de telle sorte qu’on pouvait la prendre pour un spectateur, tant la queue des chats de nos ancêtres était grosse, haute et fournie. A droite du tableau, sur un champ d’azur qui déguisait imparfaitement la pourriture du bois, les passants lisaient Guillaume ; et à gauche, Successeur du sieur Chevrel. Le soleil et la pluie avaient rongé la plus grande partie de l’or moulu parcimonieusement appliqué sur les lettres de cette inscription, dans laquelle les U remplaçaient les V et réciproquement, selon les lois de notre ancienne orthographe. Afin de rabattre l’orgueil de ceux qui croient que le monde devient de jour en jour plus spirituel, et que le moderne charlatanisme surpasse tout, il convient de faire observer ici que ces enseignes, dont l’étymologie semble bizarre à plus d’un négociant parisien, sont les tableaux morts de vivants tableaux à l’aide desquels nos espiègles ancêtres avaient réussi à amener les chalands dans leurs maisons. Ainsi la Truie-qui-file, le Singe-vert, etc., furent des animaux en cage dont l’adresse émerveillait les passants, et dont l’éducation prouvait la patience de l’industriel au quinzième siècle. De semblables curiosités enrichissaient plus vite leurs heureux possesseurs que les Providence, les Bonne-foi, les Grâce-de-Dieu et les Décollation de saint Jean-Baptiste qui se voient encore rue Saint-Denis. Cependant l’inconnu ne restait certes pas là pour admirer ce chat, qu’un moment d’attention suffisait à graver dans la mémoire. Ce jeune homme avait aussi ses singularités. Son manteau, plissé dans le goût des draperies antiques, laissait voir une élégante chaussure, d’autant plus remarquable au milieu de la boue parisienne, qu’il portait des bas de soie blancs dont les mouchetures attestaient son impatience. Il sortait sans doute d’une noce ou d’un bal car à cette heure matinale il tenait à la main des gants blancs et les boucles de ses cheveux noirs défrisés éparpillées sur ses épaules indiquaient une coiffure à la Caracalla, mise à la mode autant par l’Ecole de David que par cet engouement pour les formes grecques et romaines qui marqua les premières années de ce siècle. Malgré le bruit que faisaient quelques maraîchers attardés passant au galop pour se rendre à la grande halle, cette rue si agitée avait alors un calme dont la magie n’est connue que de ceux qui ont erré dans Paris désert, à ces heures où son tapage, un moment apaisé, renaît et s’entend dans le lointain comme la grande voix de la mer. Cet étrange jeune homme devait être aussi curieux pour les commerçants du Chat-qui-pelote, que le Chat-qui-pelote l’était pour lui. Une cravate éblouissante de blancheur rendait sa figure tourmentée encore plus pâle qu’elle ne l’était réellement. Le feu tour à tour sombre et pétillant que jetaient ses yeux noirs s’harmoniait avec les contours bizarres de son visage, avec sa bouche large et sinueuse qui se contractait en souriant. Son front, ridé par une contrariété violente, avait quelque chose de fatal. Le front n’est-il pas ce qui se trouve de plus prophétique en l’homme ? Quand celui de l’inconnu exprimait la passion, les plis qui s’y formaient causaient une sorte d’effroi par la vigueur avec laquelle ils se prononçaient ; mais lorsqu’il reprenait son calme, si facile à troubler, il y respirait une grâce lumineuse qui rendait attrayante cette physionomie où la joie, la douleur, l’amour, la colère, le dédain éclataient d’une manière si communicative que l’homme le plus froid en devait être impressionné. Cet inconnu se dépitait si bien au moment où l’on ouvrit précipitamment la lucarne du grenier, qu’il n’y vit pas apparaître trois joyeuses figures rondelettes, blanches, roses, mais aussi communes que le sont les figures du Commerce sculptées sur certains monuments. Ces trois faces, encadrées par la lucarne, rappelaient les têtes d’anges bouffis semés dans les nuages qui accompagnent le Père éternel. Les apprentis respirèrent les émanations de la rue avec une avidité qui démontrait combien l’atmosphère de leur grenier était chaude et méphitique. Après avoir indiqué ce singulier factionnaire, le commis qui paraissait être le plus jovial disparut et revint en tenant à la main un instrument dont le métal inflexible a été récemment remplacé par un cuir souple ; puis tous prirent une expression malicieuse en regardant le badaud qu’ils aspergèrent d’une pluie fine et blanchâtre dont le parfum prouvait que les trois mentons venaient d’être rasés. Elevés sur la pointe de leurs pieds, et réfugiés au fond de leur grenier pour jouir de la colère de leur victime, les commis cessèrent de rire en voyant l’insouciant dédain avec lequel le jeune homme secoua son manteau, et le profond mépris que peignit sa figure quand il leva les yeux sur la lucarne vide. En ce moment, une main blanche et délicate fit remonter vers l’imposte la partie inférieure d’une des grossières croisées du troisième étage, au moyen de ces coulisses dont le tourniquet laisse souvent tomber à l’improviste le lourd vitrage qu’il doit retenir. Le passant fut alors récompensé de sa longue attente. La figure d’une jeune fille, fraîche comme un de ces blancs calices qui fleurissent au sein des eaux, se montra couronnée d’une ruche en mousseline froissée qui donnait à sa tête un air d’innocence admirable. Quoique couverts d’une étoffe brune, son cou, ses épaules s’apercevaient, grâce à de légers interstices ménagés par les mouvements du sommeil. Aucune expression de contrainte n’altérait ni l’ingénuité de ce visage, ni le calme de ces yeux immortalisés par avance dans les sublimes compositions de Raphaël : c’était la même grâce, la même tranquillité de ces vierges devenues proverbiales. Il existait un charmant contraste produit par la jeunesse des joues de cette figure, sur laquelle le sommeil avait comme mis en relief une surabondance de vie, et par la vieillesse de cette fenêtre massive aux contours grossiers, dont l’appui était noir. Semblable à ces fleurs de jour qui n’ont pas encore au matin déplié leur tunique roulée par le froid des nuits, la jeune fille, à peine éveillée, laissa errer ses yeux bleus sur les toits voisins et regarda le ciel ; puis, par une sorte d’habitude, elle les baissa sur les sombres régions de la rue, où ils rencontrèrent aussitôt ceux de son adorateur. La coquetterie la fit sans doute souffrir d’être vue en déshabillé, elle se retira vivement en arrière, le tourniquet tout usé tourna, la croisée redescendit avec cette rapidité qui, de nos jours, a valu un nom odieux à cette naïve invention de nos ancêtres, et la vision disparut. Il semblait à ce jeune homme que la plus brillante des étoiles du matin avait été soudain cachée par un nuage.\\n\\nPendant ces petits événements, les lourds volets intérieurs qui défendaient le léger vitrage de la boutique du Chat-qui-pelote avaient été enlevés comme par magie. La vieille porte à heurtoir fut repliée sur le mur intérieur de la maison par un serviteur vraisemblablement contemporain de l’enseigne, qui d’une main tremblante y attacha le morceau de drap carré sur lequel était brodé en soie jaune le nom de Guillaume, successeur de Chevrel. Il eût été difficile à plus d’un passant de deviner le genre de commerce de monsieur Guillaume. A travers les gros barreaux de fer qui protégeaient extérieurement sa boutique, à peine y apercevait-on des paquets enveloppés de toile brune aussi nombreux que des harengs quand ils traversent l’Océan. Malgré l’apparente simplicité de cette gothique façade, monsieur Guillaume était, de tous les marchands drapiers de Paris, celui dont les magasins se trouvaient toujours le mieux fournis, dont les relations avaient le plus d’étendue, et dont la probité commerciale était la plus exacte. Si quelques-uns de ses confrères avaient conclu des marchés avec le gouvernement, sans avoir la quantité de drap voulue, il était toujours prêt à la leur livrer, quelque considérable que fût le nombre de pièces soumissionnées. Le rusé négociant connaissait mille manières de s’attribuer le plus fort bénéfice sans se trouver obligé, comme eux, de courir chez des protecteurs, y faire des bassesses ou de riches présents. Si les confrères ne pouvaient le payer qu’en excellentes traites un peu longues, il indiquait son notaire comme un homme accommodant ; et savait encore tirer une seconde mouture du sac, grâce à cet expédient qui faisait dire proverbialement aux négociants de la rue Saint-Denis : — Dieu vous garde du notaire de monsieur Guillaume ! pour désigner un escompte onéreux. Le vieux négociant se trouva debout comme par miracle, sur le seuil de sa boutique, au moment où le domestique se retira. Monsieur Guillaume regarda la rue Saint-Denis, les boutiques voisines et le temps, comme un homme qui débarque au Havre et revoit la France après un long voyage. Bien convaincu que rien n’avait changé pendant son sommeil, il aperçut alors le passant en faction, qui, de son côté, contemplait le patriarche de la draperie, comme Humboldt dut examiner le premier gymnote électrique qu’il vit en Amérique. Monsieur Guillaume portait de larges culottes de velours noir, des bas chinés, et des souliers carrés à boucles d’argent. Son habit à pans carrés, à basques carrées, à collet carré, enveloppait son corps légèrement voûté d’un drap verdâtre garni de grands boutons en métal blanc mais rougis par l’usage. Ses cheveux gris étaient si exactement aplatis et peignés sur son crâne jaune, qu’ils le faisaient ressembler à un champ sillonné. Ses petits yeux verts, percés comme avec une vrille, flamboyaient sous deux arcs marqués d’une faible rougeur à défaut de sourcils. Les inquiétudes avaient tracé sur son front des rides horizontales aussi nombreuses que les plis de son habit. Cette figure blême annonçait la patience, la sagesse commerciale, et l’espèce de cupidité rusée que réclament les affaires. A cette époque on voyait moins rarement qu’aujourd’hui de ces vieilles familles où se conservaient, comme de précieuses traditions, les moeurs, les costumes caractéristiques de leurs professions, et restées au milieu de la civilisation nouvelle comme ces débris antédiluviens retrouvés par Cuvier dans les carrières. Le chef de la famille Guillaume était un de ces notables gardiens des anciens usages : on le surprenait à regretter le Prévôt des Marchands, et jamais il ne parlait d’un jugement du tribunal de commerce sans le nommer la sentence des consuls. C’était sans doute en vertu de ces coutumes que, levé le premier de sa maison, il attendait de pied ferme l’arrivée de ses trois commis, pour les gourmander en cas de retard. Ces jeunes disciples de Mercure ne connaissaient rien de plus redoutable que l’activité silencieuse avec laquelle le patron scrutait leurs visages et leurs mouvements, le lundi matin, en y recherchant les preuves ou les traces de leurs escapades. Mais, en ce moment, le vieux drapier ne fit aucune attention à ses apprentis. Il était occupé à chercher le motif de la sollicitude avec laquelle le jeune homme en bas de soie et en manteau portait alternativement les yeux sur son enseigne et sur les profondeurs de son magasin. Le jour, devenu plus éclatant, permettait d’y apercevoir le bureau grillagé, entouré de rideaux en vieille soie verte, où se tenaient les livres immenses, oracles muets de la maison. Le trop curieux étranger semblait convoiter ce petit local, y prendre le plan d’une salle à manger latérale, éclairée par un vitrage pratiqué dans le plafond, et d’où la famille réunie devait facilement voir, pendant ses repas, les plus légers accidents qui pouvaient arriver sur le seuil de la boutique. Un si grand amour pour son logis paraissait suspect à un négociant qui avait subi le régime de la Terreur. Monsieur Guillaume pensait donc assez naturellement que cette figure sinistre en voulait à la caisse du Chat-qui-pelote. Après avoir discrètement joui du duel muet qui avait lieu entre son patron et l’inconnu, le plus âgé des commis hasarda de se placer sur la dalle où était monsieur Guillaume, en voyant le jeune homme contempler à la dérobée les croisées du troisième. Il fit deux pas dans la rue, leva la tête, et crut avoir aperçu mademoiselle Augustine Guillaume qui se retirait avec précipitation. Mécontent de la perspicacité de son premier commis, le drapier lui lança un regard de travers ; mais tout à coup les craintes mutuelles que la présence de ce passant excitait dans l’âme du marchand et de l’amoureux commis se calmèrent. L’inconnu héla un fiacre qui se rendait à une place voisine, et y monta rapidement en affectant une trompeuse indifférence. Ce départ mit un certain baume dans le coeur des autres commis, assez inquiets de retrouver la victime de leur plaisanterie.\\n\\n— Hé bien, messieurs, qu’avez-vous donc à rester là , les bras croisés ? dit monsieur Guillaume à ses trois néophytes. Mais autrefois, sarpejeu ! quand j’étais chez le sieur Chevrel, j’avais déjà visité plus de deux pièces de drap.\\n\\n— Il faisait donc jour de meilleure heure, dit le second commis que cette tâche concernait.\\n\\nLe vieux négociant ne put s’empêcher de sourire. Quoique deux de ces trois jeunes gens, confiés à ses soins par leurs pères, riches manufacturiers de Louviers et de Sedan, n’eussent qu’à demander cent mille francs pour les avoir, le jour où ils seraient en âge de s’établir, Guillaume croyait de son devoir de les tenir sous la férule d’un antique despotisme inconnu de nos jours dans les brillants magasins modernes dont les commis veulent être riches à trente ans : il les faisait travailler comme des nègres. A eux trois, ces commis suffisaient à une besogne qui aurait mis sur les dents dix de ces employés dont le sybaritisme enfle aujourd’hui les colonnes du budget. Aucun bruit ne troublait la paix de cette maison solennelle, où les gonds semblaient toujours huilés, et dont le moindre meuble avait cette propreté respectable qui annonce un ordre et une économie sévères. Souvent, le plus espiègle des commis s’était amusé à écrire sur le fromage de Gruyère qu’on leur abandonnait au déjeuner, et qu’ils se plaisaient à respecter, la date de sa réception primitive. Cette malice et quelques autres semblables faisaient parfois sourire la plus jeune des deux filles de Guillaume, la jolie vierge qui venait d’apparaître au passant enchanté. Quoique chacun des apprentis, et même le plus ancien, payât une forte pension, aucun d’eux n’eût été assez hardi pour rester à la table du patron au moment où le dessert y était servi. Lorsque madame Guillaume parlait d’accommoder la salade, ces pauvres jeunes gens tremblaient en songeant avec quelle parcimonie sa prudente main savait y épancher l’huile. Il ne fallait pas qu’ils s’avisassent de passer une nuit dehors, sans avoir donné long-temps à l’avance un motif plausible à cette irrégularité. Chaque dimanche, et à tour de rôle, deux commis accompagnaient la famille Guillaume à la messe de Saint-Leu et aux vêpres. Mesdemoiselles Virginie et Augustine, modestement vêtues d’indienne, prenaient chacune le bras d’un commis et marchaient en avant, sous les yeux perçants de leur mère, qui fermait ce petit cortége domestique avec son mari accoutumé par elle à porter deux gros paroissiens reliés en maroquin noir. Le second commis n’avait pas d’appointements. Quant à celui que douze ans de persévérance et de discrétion initiaient aux secrets de la maison, il recevait huit cents francs en récompense de ses labeurs. A certaines fêtes de famille, il était gratifié de quelques cadeaux auxquels la main sèche et ridée de madame Guillaume donnait seule du prix : des bourses en filet, qu’elle avait soin d’emplir de coton pour faire valoir leurs dessins à jour ; des bretelles fortement conditionnées, ou des paires de bas de soie bien lourdes. Quelquefois, mais rarement, ce premier ministre était admis à partager les plaisirs de la famille soit quand elle allait à la campagne, soit quand après des mois d’attente elle se décidait à user de son droit à demander, en louant une loge, une pièce à laquelle Paris ne pensait plus. Quant aux deux autres commis, la barrière de respect qui séparait jadis un maître drapier de ses apprentis était placée si fortement entre eux et le vieux négociant, qu’il leur eût été plus facile de voler une pièce de drap que de déranger cette auguste étiquette. Cette réserve peut paraître ridicule aujourd’hui. Néanmoins, ces vieilles maisons étaient des écoles de moeurs et de probité. Les maîtres adoptaient leurs apprentis. Le linge d’un jeune homme était soigné, réparé, quelquefois renouvelé par la maîtresse de la maison. Un commis tombait-il malade, il devenait l’objet de soins vraiment maternels. En cas de danger, le patron prodiguait son argent pour appeler les plus célèbres docteurs ; car il ne répondait pas seulement des moeurs et du savoir de ces jeunes gens à leurs parents. Si l’un d’eux, honorable par le caractère, éprouvait quelque désastre, ces vieux négociants savaient apprécier l’intelligence qu’ils avaient développée, et n’hésitaient pas à confier le bonheur de leurs filles à celui auquel ils avaient pendant long-temps confié leurs fortunes. Guillaume était un de ces hommes antiques, et s’il en avait les ridicules, il en avait toutes les qualités. Aussi Joseph Lebas, son premier commis, orphelin et sans fortune, était-il, dans son idée, le futur époux de Virginie sa fille aînée. Mais Joseph ne partageait point les pensées symétriques de son patron, qui, pour un empire, n’aurait pas marié sa seconde fille avant la première. L’infortuné commis se sentait le coeur entièrement pris pour mademoiselle Augustine la cadette. Afin de justifier cette passion, qui avait grandi secrètement, il est nécessaire de pénétrer plus avant dans les ressorts du gouvernement absolu qui régissait la maison du vieux marchand drapier.\\n\\nGuillaume avait deux filles. L’aînée, mademoiselle Virginie, était tout le portrait de sa mère. Madame Guillaume, fille du sieur Chevrel, se tenait si droite sur la banquette de son comptoir, que plus d’une fois elle avait entendu des plaisants parier qu’elle y était empalée. Sa figure maigre et longue trahissait une dévotion outrée. Sans grâces et sans manières aimables, madame Guillaume ornait habituellement sa tête presque sexagénaire d’un bonnet dont la forme était invariable et garni de barbes comme celui d’une veuve. Tout le voisinage l’appelait la soeur tourière. Sa parole était brève, et ses gestes avaient quelque chose des mouvements saccadés d’un télégraphe. Son oeil, clair comme celui d’un chat, semblait en vouloir à tout le monde de ce qu’elle était laide. Mademoiselle Virginie, élevée comme sa jeune soeur sous les lois despotiques de leur mère, avait atteint l’âge de vingt-huit ans. La jeunesse atténuait l’air disgracieux que sa ressemblance avec sa mère donnait parfois à sa figure ; mais la rigueur maternelle l’avait dotée de deux grandes qualités qui pouvaient tout contre-balancer : elle était douce et patiente. Mademoiselle Augustine, à peine âgée de dix-huit ans, ne ressemblait ni à son père ni à sa mère. Elle était de ces filles qui, par l’absence de tout lien physique avec leurs parents, font croire à ce dicton de prude : Dieu donne les enfants. Augustine était petite, ou, pour la mieux peindre, mignonne. Gracieuse et pleine de candeur, un homme du monde n’aurait pu reprocher à cette charmante créature que des gestes mesquins ou certaines attitudes communes, et parfois de la gêne. Sa figure silencieuse et immobile respirait cette mélancolie passagère qui s’empare de toutes les jeunes filles trop faibles pour oser résister aux volontés d’une mère. Toujours modestement vêtues, les deux soeurs ne pouvaient satisfaire la coquetterie innée chez la femme que par un luxe de propreté qui leur allait à merveille et les mettait en harmonie avec ces comptoirs luisants, avec ces rayons sur lesquels le vieux domestique ne souffrait pas un grain de poussière, avec la simplicité antique de tout ce qui se voyait autour d’elles. Obligées par leur genre de vie à chercher des éléments de bonheur dans des travaux obstinés, Augustine et Virginie n’avaient donné jusqu’alors que du contentement à leur mère, qui s’applaudissait secrètement de la perfection du caractère de ses deux filles. Il est facile d’imaginer les résultats de l’éducation qu’elles avaient reçue. Elevées pour le commerce, habituées à n’entendre que des raisonnements et des calculs tristement mercantiles, n’ayant étudié que la grammaire, la tenue des livres, un peu d’histoire juive, l’histoire de France dans Le Ragois, et ne lisant que les auteurs dont la lecture leur était permise par leur mère, leurs idées n’avaient pas pris beaucoup d’étendue : elles savaient parfaitement tenir un ménage, elles connaissaient le prix des choses, elles appréciaient les difficultés que l’on éprouve à amasser l’argent, elles étaient économes et portaient un grand respect aux qualités du négociant. Malgré la fortune de leur père, elles étaient aussi habiles à faire des reprises qu’à festonner ; souvent leur mère parlait de leur apprendre la cuisine afin qu’elles sussent bien ordonner un dîner, et pussent gronder une cuisinière en connaissance de cause. Ignorant les plaisirs du monde et voyant comment s’écoulait la vie exemplaire de leurs parents, elles ne jetaient que bien rarement leurs regards au delà de l’enceinte de cette vieille maison patrimoniale qui, pour leur mère, était l’univers. Les réunions occasionnées par les solennités de famille formaient tout l’avenir de leurs joies terrestres. Quand le grand salon situé au second étage devait recevoir madame Roguin, une demoiselle Chevrel, de quinze ans moins âgée que sa cousine et qui portait des diamants ; le jeune Rabourdin, sous-chef aux Finances ; monsieur César Birotteau, riche parfumeur, et sa femme appelée madame César ; monsieur Camusot, le plus riche négociant en soieries de la rue des Bourdonnais ; deux ou trois vieux banquiers, et des femmes irréprochables ; les apprêts nécessités par la manière dont l’argenterie, les porcelaines de Saxe, les bougies, les cristaux étaient empaquetés faisaient une diversion à la vie monotone de ces trois femmes qui allaient et venaient, en se donnant autant de mouvement que des religieuses pour la réception d’un évêque. Puis quand, le soir, fatiguées toutes trois d’avoir essuyé, frotté, déballé, mis en place les ornements de la fête, les deux jeunes filles aidaient leur mère à se coucher, madame Guillaume leur disait :\\n\\n— Nous n’avons rien fait aujourd’hui, mes enfants !\\n\\nLorsque, dans ces assemblées solennelles, la soeur tourière permettait de danser en confinant les parties de boston, de whist et de trictrac dans sa chambre à coucher, cette concession était comptée parmi les félicités les plus inespérées, et causait un bonheur égal à celui d’aller à deux ou trois grands bals où Guillaume menait ses filles à l’époque du carnaval. Enfin, une fois par an, l’honnête drapier donnait une fête pour laquelle rien n’était épargné. Quelque riches et élégantes que fussent les personnes invitées, elles se gardaient bien d’y manquer ; car les maisons les plus considérables de la place avaient recours à l’immense crédit, à la fortune ou à la vieille expérience de monsieur Guillaume. Mais les deux filles de ce digne négociant ne profitaient pas autant qu’on pourrait le supposer des enseignements que le monde offre à de jeunes âmes. Elles apportaient dans ces réunions, inscrites d’ailleurs sur le carnet d’échéances de la maison, des parures dont la mesquinerie les faisait rougir. Leur manière de danser n’avait rien de remarquable, et la surveillance maternelle ne leur permettait pas de soutenir la conversation autrement que par Oui et Non avec leurs cavaliers. Puis la loi de la vieille enseigne du Chat-qui-pelote leur ordonnait d’être rentrées à onze heures, moment où les bals et les fêtes commencent à s’animer. Ainsi leurs plaisirs, en apparence assez conformes à la fortune de leur père, devenaient souvent insipides par des circonstances qui tenaient aux habitudes et aux principes de cette famille. Quant à leur vie habituelle, une seule observation achèvera de la peindre. Madame Guillaume exigeait que ses deux filles fussent habillées de grand matin, qu’elles descendissent tous les jours à la même heure, et soumettait leurs occupations à une régularité monastique. Cependant Augustine avait reçu du hasard une âme assez élevée pour sentir le vide de cette existence. Parfois ses yeux bleus se relevaient comme pour interroger les profondeurs de cet escalier sombre et de ces magasins humides. Après avoir sondé ce silence de cloître, elle semblait écouter de loin de confuses révélations de cette vie passionnée qui met les sentiments à un plus haut prix que les choses. En ces moments son visage se colorait, ses mains inactives laissaient tomber la blanche mousseline sur le chêne poli du comptoir, et bientôt sa mère lui disait d’une voix qui restait toujours aigre même dans les tons les plus doux :\\n\\n— Augustine ! à quoi pensez-vous donc, mon bijou ?\\n\\nPeut-être Hippolyte comte de Douglas et le Comte de Comminges, deux romans trouvés par Augustine dans l’armoire d’une cuisinière récemment renvoyée par madame Guillaume, contribuèrent-ils à développer les idées de cette jeune fille qui les avait furtivement dévorés pendant les longues nuits de l’hiver précédent. Les expressions de désir vague, la voix douce, la peau de jasmin et les yeux bleus d’Augustine avaient donc allumé dans l’âme du pauvre Lebas un amour aussi violent que respectueux. Par un caprice facile à comprendre, Augustine ne se sentait aucun goût pour l’orphelin : peut-être était-ce parce qu’elle ne se savait pas aimée. En revanche, les longues jambes, les cheveux châtains, les grosses mains et l’encolure vigoureuse du premier commis avaient trouvé une secrète admiratrice dans mademoiselle Virginie, qui, malgré ses cinquante mille écus de dot, n’était demandée en mariage par personne. Rien de plus naturel que ces deux passions inverses nées dans le silence de ces comptoirs obscurs comme fleurissent des violettes dans la profondeur d’un bois. La muette et constante contemplation qui réunissait les yeux de ces jeunes gens par un besoin violent de distraction au milieu de travaux obstinés et d’une paix religieuse, devait tôt ou tard exciter des sentiments d’amour. L’habitude de voir une figure y fait découvrir insensiblement les qualités de l’âme, et finit par en effacer les défauts.\\n\\n— Au train dont y va cet homme, nos filles ne tarderont pas à se mettre à genoux devant un prétendu ! se dit monsieur Guillaume en lisant le premier décret par lequel Napoléon anticipa sur les classes de conscrits.\\n\\nDès ce jour, désespéré de voir sa fille aînée se faner, le vieux marchand se souvint d’avoir épousé mademoiselle Chevrel à peu près dans la situation où se trouvaient Joseph Lebas et Virginie. Quelle belle affaire que de marier sa fille et d’acquitter une dette sacrée, en rendant à un orphelin le bienfait qu’il avait reçu jadis de son prédécesseur dans les mêmes circonstances ! Agé de trente-trois ans, Joseph Lebas pensait aux obstacles que quinze ans de différence mettaient entre Augustine et lui. Trop perspicace d’ailleurs pour ne pas deviner les desseins de monsieur Guillaume, il en connaissait assez les principes inexorables pour savoir que jamais la cadette ne se marierait avant l’aînée. Le pauvre commis, dont le coeur était aussi excellent que ses jambes étaient longues et son buste épais, souffrait donc en silence.\\n\\nTel était l’état des choses dans cette petite république, qui, au milieu de la rue Saint-Denis, ressemblait assez à une succursale de la Trappe. Mais pour rendre un compte exact des événements extérieurs comme des sentiments, il est nécessaire de remonter à quelques mois avant la scène par laquelle commence cette histoire. A la nuit tombante, un jeune homme passant devant l’obscure boutique du Chat-qui-pelote y était resté un moment en contemplation à l’aspect d’un tableau qui aurait arrêté tous les peintres du monde. Le magasin, n’étant pas encore éclairé, formait un plan noir au fond duquel se voyait la salle à manger du marchand. Une lampe astrale y répandait ce jour jaune qui donne tant de grâce aux tableaux de l’école hollandaise. Le linge blanc, l’argenterie, les cristaux formaient de brillants accessoires qu’embellissaient encore de vives oppositions entre l’ombre et la lumière. La figure du père de famille et celle de sa femme, les visages des commis et les formes pures d’Augustine, à deux pas de laquelle se tenait une grosse fille joufflue, composaient un groupe si curieux ; ces têtes étaient si originales, et chaque caractère avait une expression si franche ; on devinait si bien la paix, le silence et la modeste vie de cette famille, que, pour un artiste accoutumé à exprimer la nature, il y avait quelque chose de désespérant à vouloir rendre cette scène fortuite. Ce passant était un jeune peintre, qui, sept ans auparavant, avait remporté le grand prix de peinture. Il revenait de Rome. Son âme nourrie de poésie, ses yeux rassasiés de Raphaël et de Michel-Ange, avaient soif de la nature vraie, après une longue habitation du pays pompeux où l’art a jeté partout son grandiose. Faux ou juste, tel était son sentiment personnel. Abandonné long-temps à la fougue des passions italiennes, son coeur demandait une de ces vierges modestes et recueillies que, malheureusement, il n’avait su trouver qu’en peinture à Rome. De l’enthousiasme imprimé à son âme exaltée par le tableau naturel qu’il contemplait, il passa naturellement à une profonde admiration pour la figure principale : Augustine paraissait pensive et ne mangeait point ; par une disposition de la lampe dont la lumière tombait entièrement sur son visage, son buste semblait se mouvoir dans un cercle de feu qui détachait plus vivement les contours de sa tête et l’illuminait d’une manière quasi surnaturelle. L’artiste la compara involontairement à un ange exilé qui se souvient du ciel. Une sensation presque inconnue, un amour limpide et bouillonnant inonda son coeur. Après être demeuré pendant un moment comme écrasé sous le poids de ses idées, il s’arracha à son bonheur, rentra chez lui, ne mangea pas, ne dormit point. Le lendemain, il entra dans son atelier pour n’en sortir qu’après avoir déposé sur une toile la magie de cette scène dont le souvenir l’avait en quelque sorte fanatisé. Sa félicité fut incomplète tant qu’il ne posséda pas un fidèle portrait de son idole. Il passa plusieurs fois devant la maison du Chat-qui-pelote ; il osa même y entrer une ou deux fois sous le masque d’un déguisement, afin de voir de plus près la ravissante créature que madame Guillaume couvrait de son aile. Pendant huit mois entiers, adonné à son amour, à ses pinceaux, il resta invisible pour ses amis les plus intimes, oubliant le monde, la poésie, le théâtre, la musique, et ses plus chères habitudes. Un matin, Girodet força toutes ces consignes que les artistes connaissent et savent éluder, parvint à lui et le réveilla par cette demande :\\n\\n— Que mettras-tu au Salon ?\\n\\nL’artiste saisit la main de son ami, l’entraîne à son atelier, découvre un petit tableau de chevalet et un portrait. Après une lente et avide contemplation des deux chefs-d’oeuvre, Girodet saute au cou de son camarade et l’embrasse, sans trouver de paroles. Ses émotions ne pouvaient se rendre que comme il les sentait, d’âme à âme.\\n\\n— Tu es amoureux ? dit Girodet.\\n\\nTous deux savaient que les plus beaux portraits de Titien, de Raphaël et de Léonard de Vinci sont dus à des sentiments exaltés, qui, sous diverses conditions, engendrent d’ailleurs tous les chefs-d’oeuvre. Pour toute réponse, le jeune artiste inclina la tête.\\n\\n— Es-tu heureux de pouvoir être amoureux ici, en revenant d’Italie ! Je ne te conseille pas de mettre de telles oeuvres au Salon, ajouta le grand peintre. Vois-tu, ces deux tableaux n’y seraient pas sentis. Ces couleurs vraies, ce travail prodigieux ne peuvent pas encore être appréciés, le public n’est plus accoutumé à tant de profondeur. Les tableaux que nous peignons, mon bon ami, sont des écrans, des paravents. Tiens, faisons plutôt des vers, et traduisons les Anciens ! il y a plus de gloire à en attendre, que de nos malheureuses toiles.\\n\\nMalgré cet avis charitable, les deux toiles furent exposées. La scène d’intérieur fit une révolution dans la peinture. Elle donna naissance à ces tableaux de genre dont la prodigieuse quantité importée à toutes nos expositions, pourrait faire croire qu’ils s’obtiennent par des procédés purement mécaniques. Quant au portrait, il est peu d’artistes qui ne gardent le souvenir de cette toile vivante à laquelle le public, quelquefois juste en masse, laissa la couronne que Girodet y plaça lui-même. Les deux tableaux furent entourés d’une foule immense. On s’y tua, comme disent les femmes. Des spéculateurs, des grands seigneurs couvrirent ces deux toiles de doubles napoléons, l’artiste refusa obstinément de les vendre, et refusa d’en faire des copies. On lui offrit une somme énorme pour les laisser graver, les marchands ne furent pas plus heureux que ne l’avaient été les amateurs. Quoique cette aventure fît du bruit dans le monde, elle n’était pas de nature à parvenir au fond de la petite Thébaïde de la rue Saint-Denis. Néanmoins, en venant faire une visite à madame Guillaume, la femme du notaire parla de l’exposition devant Augustine, qu’elle aimait beaucoup, et lui en expliqua le but. Le babil de madame Roguin inspira naturellement à Augustine le désir de voir les tableaux, et la hardiesse de demander secrètement à sa cousine de l’accompagner au Louvre. La cousine réussit dans la négociation qu’elle entama auprès de madame Guillaume, pour obtenir la permission d’arracher sa petite cousine à ses tristes travaux pendant environ deux heures. La jeune fille pénétra donc, à travers la foule, jusqu’au tableau couronné. Un frisson la fit trembler comme une feuille de bouleau, quand elle se reconnut. Elle eut peur et regarda autour d’elle pour rejoindre madame Roguin, de qui elle avait été séparée par un flot de monde. En ce moment ses yeux effrayés rencontrèrent la figure enflammée du jeune peintre. Elle se rappela tout à coup la physionomie d’un promeneur que, curieuse, elle avait souvent remarqué, en croyant que c’était un nouveau voisin.\\n\\n— Vous voyez ce que l’amour m’a fait faire, dit l’artiste à l’oreille de la timide créature qui resta tout épouvantée de ces paroles.\\n\\nElle trouva un courage surnaturel pour fendre la presse, et pour rejoindre sa cousine encore occupée à percer la masse du monde qui l’empêchait d’arriver jusqu’au tableau.\\n\\n— Vous seriez étouffée, s’écria Augustine, partons !\\n\\nMais il se rencontre, au Salon, certains moments pendant lesquels deux femmes ne sont pas toujours libres de diriger leurs pas dans les galeries. Mademoiselle Guillaume et sa cousine furent poussées à quelques pas du second tableau, par suite des mouvements irréguliers que la foule leur imprima. Le hasard voulut qu’elles eussent la facilité d’approcher ensemble de la toile illustrée par la mode, d’accord cette fois avec le talent. La femme du notaire fit une exclamation de surprise perdue dans le brouhaha et les bourdonnements de la foule ; mais Augustine pleura involontairement à l’aspect de cette merveilleuse scène. Puis, par un sentiment presque inexplicable, elle mit un doigt sur ses lèvres en apercevant à deux pas d’elle la figure extatique du jeune artiste. L’inconnu répondit par un signe de tête et désigna madame Roguin, comme un trouble-fête, afin de montrer à Augustine qu’elle était comprise. Cette pantomime jeta comme un brasier dans le corps de la pauvre fille qui se trouva criminelle, en se figurant qu’il venait de se conclure un pacte entre elle et l’artiste. Une chaleur étouffante, le continuel aspect des plus brillantes toilettes, et l’étourdissement que produisaient sur Augustine la variété des couleurs, la multitude des figures vivantes ou peintes, la profusion des cadres d’or, lui firent éprouver une espèce d’enivrement qui redoubla ses craintes. Elle se serait peut-être évanouie, si, malgré ce chaos de sensations, il ne s’était élevé au fond de son coeur une jouissance inconnue qui vivifia tout son être. Néanmoins, elle se crut sous l’empire de ce démon dont les terribles piéges lui étaient prédits par la voix tonnante des prédicateurs. Ce moment fut pour elle comme un moment de folie. Elle se vit accompagnée jusqu’à la voiture de sa cousine par ce jeune homme resplendissant de bonheur et d’amour. En proie à une irritation toute nouvelle, une ivresse qui la livrait en quelque sorte à la nature, Augustine écouta la voix éloquente de son coeur, et regarda plusieurs fois le jeune peintre en laissant paraître le trouble dont elle était saisie. Jamais l’incarnat de ses joues n’avait formé de plus vigoureux contrastes avec la blancheur de sa peau. L’artiste aperçut alors cette beauté dans toute sa fleur, cette pudeur dans toute sa gloire. Augustine éprouva une sorte de joie mêlée de terreur, en pensant que sa présence causait la félicité de celui dont le nom était sur toutes les lèvres, dont le talent donnait l’immortalité à de passagères images. Elle était aimée ! il lui était impossible d’en douter. Quand elle ne vit plus l’artiste, elle entendit encore retentir dans son coeur ces paroles simples : « Vous voyez ce que l’amour m’a fait faire. » Et les palpitations devenues plus profondes lui semblèrent une douleur, tant son sang plus ardent réveilla dans son corps de puissances inconnues. Elle feignit d’avoir un grand mal de tête pour éviter de répondre aux questions de sa cousine relativement aux tableaux ; mais, au retour, madame Roguin ne put s’empêcher de parler à madame Guillaume de la célébrité obtenue par le Chat-qui-pelote, et Augustine trembla de tous ses membres en entendant dire à sa mère qu’elle irait au Salon pour y voir sa maison. La jeune fille insista de nouveau sur sa souffrance, et obtint la permission d’aller se coucher.\\n\\n— Voilà ce qu’on gagne à tous ces spectacles, s’écria monsieur Guillaume, des maux de tête. Est-ce donc bien amusant de voir en peinture ce qu’on rencontre tous les jours dans notre rue ! Ne me parlez pas de ces artistes qui sont, comme vos auteurs, des meure-de-faim. Que diable ont-ils besoin de prendre ma maison pour la vilipender dans leurs tableaux ?\\n\\n— Cela pourra nous faire vendre quelques aunes de drap de plus, dit\\n\\nJoseph Lebas.\\n\\n\\n\\nCette observation n’empêcha pas que les arts et la pensée ne fussent condamnés encore une fois au tribunal du Négoce. Comme on doit bien le penser, ces discours ne donnèrent pas grand espoir à Augustine. Elle eut toute la nuit pour se livrer à la première méditation de l’amour. Les événements de cette journée furent comme un songe qu’elle se plut à reproduire dans sa pensée Elle s’initia aux craintes, aux espérances, aux remords, à toutes ces ondulations de sentiment qui devaient bercer un coeur simple et timide comme le sien. Quel vide elle reconnut dans cette noire maison, et quel trésor elle trouva dans son âme ! Etre la femme d’un homme de talent, partager sa gloire ! Quels ravages cette idée ne devait-elle pas faire au coeur d’une enfant élevée au sein de cette famille ! Quelle espérance ne devait-elle pas éveiller chez une jeune personne qui, nourrie jusqu’alors de principes vulgaires, avait désiré une vie élégante ! Un rayon de soleil était tombé dans cette prison. Augustine aima tout à coup. En elle tant de sentiments étaient flattés à la fois, qu’elle succomba sans rien calculer. A dix-huit ans, l’amour ne jette-t-il pas son prisme entre le monde et les yeux d’une jeune fille ? Incapable de deviner les rudes chocs qui résultent de l’alliance d’une femme aimante avec un homme d’imagination, elle crut être appelée à faire le bonheur de celui-ci, sans apercevoir aucune disparate entre elle et lui. Pour elle, le présent fut tout l’avenir. Quand le lendemain son père et sa mère revinrent du Salon, leurs figures attristées annoncèrent quelque désappointement. D’abord, les deux tableaux avaient été retirés par le peintre ; puis, madame Guillaume avait perdu son châle de cachemire. Apprendre que les tableaux venaient de disparaître après sa visite au Salon fut pour Augustine la révélation d’une délicatesse de sentiment que les femmes savent toujours apprécier, même instinctivement.\\n\\nLe matin où, rentrant d’un bal, Théodore de Sommervieux, tel était le nom que la renommée avait apporté dans le coeur d’Augustine, fut aspergé par les commis du Chat-qui-pelote pendant qu’il attendait l’apparition de sa naïve amie, qui ne le savait certes pas là , les deux amants se voyaient pour la quatrième fois seulement depuis la scène du Salon. Les obstacles que le régime de la maison Guillaume opposait au caractère fougueux de l’artiste, donnaient à sa passion pour Augustine une violence facile à concevoir. Comment aborder une jeune fille assise dans un comptoir entre deux femmes telles que mademoiselle Virginie et madame Guillaume ? Comment correspondre avec elle, quand sa mère ne la quittait jamais ? Habile, comme tous les amants, à se forger des malheurs, Théodore se créait un rival dans l’un des commis, et mettait les autres dans les intérêts de son rival. S’il échappait à tant d’Argus, il se voyait échouant sous les yeux sévères du vieux négociant ou de madame Guillaume. Partout des barrières, partout le désespoir ! La violence même de sa passion empêchait le jeune peintre de trouver ces expédients ingénieux qui, chez les prisonniers comme chez les amants, semblent être le dernier effort de la raison échauffée par un sauvage besoin de liberté ou par le feu de l’amour. Théodore tournait alors dans le quartier avec l’activité d’un fou, comme si le mouvement pouvait lui suggérer des ruses. Après s’être bien tourmenté l’imagination, il inventa de gagner à prix d’or la servante joufflue. Quelques lettres furent donc échangées de loin en loin pendant la quinzaine qui suivit la malencontreuse matinée où monsieur Guillaume et Théodore s’étaient si bien examinés.\\n\\nEn ce moment, les deux jeunes gens étaient convenus de se voir à une certaine heure du jour et le dimanche, à Saint-Leu, pendant la messe et les vêpres. Augustine avait envoyé à son cher Théodore la liste des parents et des amis de la famille, chez lesquels le jeune peintre tâcha d’avoir accès afin d’intéresser à ses amoureuses pensées, s’il était possible, une de ces âmes occupées d’argent, de commerce, et auxquelles une passion véritable devait sembler la spéculation la plus monstrueuse, une spéculation inouïe. D’ailleurs, rien ne changea dans les habitudes du Chat-qui-pelote. Si Augustine fut distraite, si, contre toute espèce d’obéissance aux lois de la charte domestique, elle monta à sa chambre pour y aller, grâce à un pot de fleurs, établir des signaux ; si elle soupira, si elle pensa enfin, personne, pas même sa mère, ne s’en aperçut. Cette circonstance causera quelque surprise à ceux qui auront compris l’esprit de cette maison, où une pensée entachée de poésie devait produire un contraste avec les êtres et les choses, où personne ne pouvait se permettre ni un geste, ni un regard qui ne fussent vus et analysés. Cependant rien de plus naturel : le vaisseau si tranquille qui naviguait sur la mer orageuse de la place de Paris, sous le pavillon du Chat-qui-pelote, était la proie d’une de ces tempêtes qu’on pourrait nommer équinoxiales à cause de leur retour périodique. Depuis quinze jours, les quatre hommes de l’équipage, madame Guillaume et mademoiselle Virginie s’adonnaient à ce travail excessif désigné sous le nom d’inventaire. On remuait tous les ballots et l’on vérifiait l’aunage des pièces pour s’assurer de la valeur exacte du coupon. On examinait soigneusement la carte appendue au paquet pour reconnaître en quel temps les draps avaient été achetés. On fixait le prix actuel. Toujours debout, son aune à la main, la plume derrière l’oreille, monsieur Guillaume ressemblait à un capitaine commandant la manoeuvre. Sa voix aiguë, passant par un judas pour interroger la profondeur des écoutilles du magasin d’en bas, faisait entendre ces barbares locutions du commerce, qui ne s’exprime que par énigmes : — Combien d’H-N-Z ? — Enlevé. — Que reste-t-il de Q-X ? — Deux aunes. — Quel prix ? — Cinq-cinq-trois. — Portez à trois A tout J-J, tout M-P, et le reste de V-D-O. Mille autres phrases tout aussi intelligibles ronflaient à travers les comptoirs comme des vers de la poésie moderne que des romantiques se seraient cités afin d’entretenir leur enthousiasme pour un de leurs poètes. Le soir, Guillaume, enfermé avec son commis et sa femme, soldait les comptes, portait à nouveau, écrivait aux retardataires, et dressait des factures. Tous trois préparaient ce travail immense dont le résultat tenait sur un carré de papier tellière, et prouvait à la maison Guillaume qu’il existait tant en argent, tant en marchandises, tant en traites et billets ; qu’elle ne devait pas un sou, qu’il lui était dû cent ou deux cent mille francs ; que le capital avait augmenté ; que les fermes, les maisons, les rentes allaient être ou arrondies, ou réparées, ou doublées. De là résultait la nécessité de recommencer avec plus d’ardeur que jamais à ramasser de nouveaux écus, sans qu’il vînt en tête à ces courageuses fourmis de se demander : A quoi bon ?\\n\\nA la faveur de ce tumulte annuel, l’heureuse Augustine échappait à l’investigation de ses Argus. Enfin, un samedi soir, la clôture de l’inventaire eut lieu. Les chiffres du total actif offrirent assez de zéros pour qu’en cette circonstance Guillaume levât la consigne sévère qui régnait toute l’année au dessert. Le sournois drapier se frotta les mains, et permit à ses commis de rester à sa table. A peine chacun des hommes de l’équipage achevait-il son petit verre d’une liqueur de ménage, on entendit le roulement d’une voiture. La famille alla voir Cendrillon aux Variétés, tandis que les deux derniers commis reçurent chacun un écu de six francs et la permission d’aller où bon leur semblerait, pourvu qu’ils fussent rentrés à minuit. Malgré cette débauche, le dimanche matin, le vieux marchand drapier fit sa barbe dès six heures, endossa son habit marron dont les superbes reflets lui causaient toujours le même contentement, il attacha des boucles d’or aux oreilles de son ample culotte de soie ; puis, vers sept heures, au moment où tout dormait encore dans la maison, il se dirigea vers le petit cabinet attenant à son magasin du premier étage. Le jour y venait d’une croisée armée de gros barreaux de fer, et qui donnait sur une petite cour carrée formée de murs si noirs qu’elle ressemblait assez à un puits. Le vieux négociant ouvrit lui-même ces volets garnis de tôle qu’il connaissait si bien, et releva une moitié du vitrage en le faisant glisser dans sa coulisse. L’air glacé de la cour vint rafraîchir la chaude atmosphère de ce cabinet, qui exhalait l’odeur particulière aux bureaux. Le marchand resta debout la main posée sur le bras crasseux d’un fauteuil de canne doublé de maroquin dont la couleur primitive était effacée, il semblait hésiter à s’y asseoir. Il regarda d’un air attendri le bureau à double pupitre, où la place de sa femme se trouvait ménagée, dans le côté opposé à la sienne, par une petite arcade pratiquée dans le mur. Il contempla les cartons numérotés, les ficelles, les ustensiles, les fers à marquer le drap, la caisse, objets d’une origine immémoriale, et crut se revoir devant l’ombre évoquée du sieur Chevrel. Il avança le même tabouret sur lequel il s’était jadis assis en présence de son défunt patron. Ce tabouret garni de cuir noir, et dont le crin s’échappait depuis long-temps par les coins mais sans se perdre, il le plaça d’une main tremblante au même endroit où son prédécesseur l’avait mis ; puis, dans une agitation difficile à décrire, il tira la sonnette qui correspondait au chevet du lit de Joseph Lebas. Quand ce coup décisif eut été frappé, le vieillard, pour qui ces souvenirs furent sans doute trop lourds, prit trois ou quatre lettres de change qui lui avaient été présentées, et les regarda sans les voir, quand Joseph Lebas se montra soudain.\\n\\n— Asseyez-vous là , lui dit Guillaume en lui désignant le tabouret.\\n\\nComme jamais le vieux maître-drapier n’avait fait asseoir son commis devant lui, Joseph Lebas tressaillit.\\n\\n— Que pensez-vous de ces traites ? demanda Guillaume.\\n\\n— Elles ne seront pas payées.\\n\\n— Comment ?\\n\\n— Mais j’ai su qu’avant-hier Etienne et compagnie ont fait leurs paiements en or.\\n\\n— Oh ! oh ! s’écria le drapier, il faut être bien malade pour laisser voir sa bile. Parlons d’autre chose. Joseph, l’inventaire est fini.\\n\\n— Oui, monsieur, et le dividende est un des plus beaux que vous ayez eus.\\n\\n— Ne vous servez donc pas de ces nouveaux mots ! Dites le produit, Joseph. Savez-vous, mon garçon, que c’est un peu à vous que nous devons ces résultats ! aussi, ne veux-je plus que vous ayez d’appointements. Madame Guillaume m’a donné l’idée de vous offrir un intérêt. Hein, Joseph ! Guillaume et Lebas, ces mots ne feraient-ils pas une belle raison sociale ? On pourrait mettre et compagnie pour arrondir la signature.\\n\\nLes larmes vinrent aux yeux de Joseph Lebas, qui s’efforça de les cacher.\\n\\n— Ah, monsieur Guillaume ! comment ai-je pu mériter tant de bontés ? Je n’ai fait que mon devoir. C’était déjà tant que de vous intéresser à un pauvre orph…\\n\\nIl brossait le parement de sa manche gauche avec la manche droite, et n’osait regarder le vieillard qui souriait en pensant que ce modeste jeune homme avait sans doute besoin, comme lui autrefois, d’être encouragé pour rendre l’explication complète.\\n\\n— Cependant, reprit le père de Virginie, vous ne méritez pas beaucoup cette faveur, Joseph ! Vous ne mettez pas en moi autant de confiance que j’en mets en vous. (Le commis releva brusquement la tête.)\\n\\n— Vous avez le secret de la caisse. Depuis deux ans je vous ai dit presque toutes mes affaires. Je vous ai fait voyager en fabrique. Enfin, pour vous, je n’ai rien sur le coeur. Mais vous ?… vous avez une inclination, et ne m’en avez pas touché un seul mot. (Joseph Lebas rougit.)\\n\\n— Ah ! ah ! s’écria Guillaume, vous pensiez donc tromper un vieux renard comme moi ? Moi ! à qui vous avez vu deviner la faillite Lecoq.\\n\\n— Comment, monsieur ? répondit Joseph Lebas en examinant son patron avec autant d’attention que son patron l’examinait, comment, vous sauriez qui j’aime ?\\n\\n— Je sais tout, vaurien, lui dit le respectable et rusé marchand en lui tordant le bout de l’oreille. Et je te pardonne, j’ai fait de même.\\n\\n— Et vous me l’accorderiez ?\\n\\n— Oui, avec cinquante mille écus, et je t’en laisserai autant, et nous marcherons sur nouveaux frais avec une nouvelle raison sociale. Nous brasserons encore des affaires, garçon, s’écria le vieux marchand en s’exaltant, se levant et agitant ses bras. Vois-tu, mon gendre, il n’y a que le commerce ! Ceux qui se demandent quels plaisirs on y trouve sont des imbéciles. Etre à la piste des affaires, savoir gouverner sur la place, attendre avec anxiété, comme au jeu, si les Etienne et compagnie font faillite, voir passer un régiment de la garde impériale habillé de notre drap, donner un croc en jambe au voisin, loyalement s’entend ! fabriquer à meilleur marché que les autres ; suivre une affaire qu’on ébauche, qui commence, grandit, chancelle et réussit ; connaître comme un ministre de la police tous les ressorts des maisons de commerce pour ne pas faire fausse route ; se tenir debout devant les naufrages ; avoir des amis, par correspondance, dans toutes les villes manufacturières, n’est-ce pas un jeu perpétuel, Joseph ? Mais c’est vivre, ça ! Je mourrai dans ce tracas-là , comme le vieux Chevrel, n’en prenant cependant plus qu’à mon aise. Dans la chaleur de sa plus forte improvisation, le père Guillaume n’avait presque pas regardé son commis qui pleurait à chaudes larmes.\\n\\n— Eh bien ! Joseph, mon pauvre garçon, qu’as-tu donc ?\\n\\n— Ah ! je l’aime tant, tant, monsieur Guillaume, que le coeur me manque, je crois…\\n\\n— Eh bien ! garçon, dit le marchand attendri, tu es plus heureux que tu ne crois, sarpejeu, car elle t’aime. Je le sais, moi !\\n\\nEt il cligna ses deux petits yeux verts en regardant son commis.\\n\\n— Mademoiselle Augustine, mademoiselle Augustine ! s’écria Joseph\\n\\nLebas dans son enthousiasme.\\n\\n\\n\\nIl allait s’élancer hors du cabinet, quand il se sentit arrêté par un bras de fer, et son patron stupéfait le ramena vigoureusement devant lui.\\n\\n— Qu’est-ce que fait donc Augustine dans cette affaire-là ? demanda Guillaume dont la voix glaça sur-le-champ le malheureux Joseph Lebas.\\n\\n— N’est-ce pas elle… que… j’aime ? dit le commis en balbutiant. Déconcerté de son défaut de perspicacité, Guillaume se rassit et mit sa tête pointue dans ses deux mains pour réfléchir à la bizarre position dans laquelle il se trouvait. Joseph Lebas honteux et au désespoir resta debout.\\n\\n— Joseph, reprit le négociant avec une dignité froide, je vous parlais de Virginie. L’amour ne se commande pas, je le sais. Je connais votre discrétion, nous oublierons cela. Je ne marierai jamais Augustine avant Virginie. Votre intérêt sera de dix pour cent.\\n\\nLe commis, auquel l’amour donna je ne sais quel degré de courage et d’éloquence, joignit les mains, prit la parole, parla pendant un quart d’heure à Guillaume avec tant de chaleur et de sensibilité, que la situation changea. S’il s’était agi d’une affaire commerciale, le vieux négociant aurait eu des règles fixes pour prendre une résolution ; mais, jeté à mille lieues du commerce, sur la mer des sentiments, et sans boussole, il flotta irrésolu devant un événement si original, se disait-il. Entraîné par sa bonté naturelle, il battit un peu la campagne.\\n\\n— Et, diantre, Joseph, tu n’es pas sans savoir que j’ai eu mes deux enfants à dix ans de distance ! Mademoiselle Chevrel n’était pas belle, elle n’a cependant pas à se plaindre de moi. Fais donc comme moi. Enfin, ne pleure pas, es-tu bête ? Que veux-tu ? cela s’arrangera peut-être, nous verrons. Il y a toujours moyen de se tirer d’affaire. Nous autres hommes nous ne sommes pas toujours comme des Céladons pour nos femmes. Tu m’entends ? Madame Guillaume est dévote, et… Allons, sarpejeu, mon enfant, donne ce matin le bras à Augustine pour aller à la messe.\\n\\nTelles furent les phrases jetées à l’aventure par Guillaume. La conclusion qui les terminait ravit l’amoureux commis : il songeait déjà pour mademoiselle Virginie à l’un de ses amis, quand il sortit du cabinet enfumé en serrant la main de son futur beau-père, après lui avoir dit, d’un petit air entendu, que tout s’arrangerait au mieux. « Que va penser madame Guillaume ? » Cette idée tourmenta prodigieusement le brave négociant quand il fut seul.\\n\\nAu déjeuner, madame Guillaume et Virginie, auxquelles le marchand-drapier avait laissé provisoirement ignorer son désappointement, regardèrent assez malicieusement Joseph Lebas qui resta grandement embarrassé. La pudeur du commis lui concilia l’amitié de sa belle-mère. La matrone redevint si gaie qu’elle regarda monsieur Guillaume en souriant, et se permit quelques petites plaisanteries d’un usage immémorial dans ces innocentes familles. Elle mit en question la conformité de la taille de Virginie et de celle de Joseph, pour leur demander de se mesurer. Ces niaiseries préparatoires attirèrent quelques nuages sur le front du chef de famille, et il afficha même un tel amour pour le décorum, qu’il ordonna à Augustine de prendre le bras du premier commis en allant à Saint-Leu. Madame Guillaume, étonnée de cette délicatesse masculine, honora son mari d’un signe de tête d’approbation. Le cortége partit donc de la maison dans un ordre qui ne pouvait suggérer aucune interprétation malicieuse aux voisins.\\n\\n— Ne trouvez-vous pas, mademoiselle Augustine, disait le commis en tremblant, que la femme d’un négociant qui a un bon crédit, comme monsieur Guillaume, par exemple, pourrait s’amuser un peu plus que ne s’amuse madame votre mère, pourrait porter des diamants, aller en voiture ? Oh ! moi, d’abord, si je me mariais, je voudrais avoir toute la peine, et voir ma femme heureuse. Je ne la mettrais pas dans mon comptoir. Voyez-vous, dans la draperie, les femmes n’y sont plus aussi nécessaires qu’elles l’étaient autrefois. Monsieur Guillaume a eu raison d’agir comme il a fait, et d’ailleurs c’était le goût de son épouse. Mais qu’une femme sache donner un coup de main à la comptabilité, à la correspondance, au détail, aux commandes, à son ménage, afin de ne pas rester oisive, c’est tout. A sept heures, quand la boutique serait fermée, moi je m’amuserais, j’irais au spectacle et dans le monde. Mais vous ne m’écoutez pas.\\n\\n— Si fait, monsieur Joseph. Que dites-vous de la peinture ? C’est là un bel état.\\n\\n— Oui, je connais un maître peintre en bâtiment, monsieur Lourdois, qui a des écus.\\n\\nEn devisant ainsi, la famille atteignit l’église de Saint-Leu. Là , madame Guillaume retrouva ses droits, et fit mettre, pour la première fois, Augustine à côté d’elle. Virginie prit place sur la quatrième chaise à côté de Lebas. Pendant le prône, tout alla bien entre Augustine et Théodore qui, debout derrière un pilier, priait sa madone avec ferveur ; mais au lever-Dieu, madame Guillaume s’aperçut, un peu tard, que sa fille Augustine tenait son livre de messe au rebours. Elle se disposait à la gourmander vigoureusement, quand, rabaissant son voile, elle interrompit sa lecture et se mit à regarder dans la direction qu’affectionnaient les yeux de sa fille. A l’aide de ses bésicles, elle vit le jeune artiste dont l’élégance mondaine annonçait plutôt quelque capitaine de cavalerie en congé, qu’un négociant du quartier. Il est difficile d’imaginer l’état violent dans lequel se trouva madame Guillaume, qui se flattait d’avoir parfaitement élevé ses filles, en reconnaissant dans le coeur d’Augustine un amour clandestin dont le danger lui fut exagéré par sa pruderie et par son ignorance. Elle crut sa fille gangrenée jusqu’au coeur.\\n\\n— Tenez d’abord votre livre à l’endroit, mademoiselle, dit-elle à voix basse mais en tremblant de colère. Elle arracha vivement le Paroissien accusateur, et le remit de manière à ce que les lettres fussent dans leur sens naturel.\\n\\n— N’ayez pas le malheur de lever les yeux autre part que sur vos prières, ajouta-t-elle, autrement, vous auriez affaire à moi. Après la messe, votre père et moi nous aurons à vous parler.\\n\\nCes paroles furent comme un coup de foudre pour la pauvre Augustine. Elle se sentit défaillir ; mais combattue entre la douleur qu’elle éprouvait et la crainte de faire un esclandre dans l’église, elle eut le courage de cacher ses angoisses. Cependant, il était facile de deviner l’état violent de son âme en voyant son Paroissien trembler et des larmes tomber sur chacune des pages qu’elle tournait. Au regard enflammé que lui lança madame Guillaume, l’artiste vit le péril où tombaient ses amours, et sortit, la rage dans le coeur, décidé à tout oser.\\n\\n— Allez dans votre chambre, mademoiselle ! dit madame Guillaume à sa fille en rentrant au logis ; nous vous ferons appeler ; et surtout, ne vous avisez pas d’en sortir.\\n\\nLa conférence que les deux époux eurent ensemble fut si secrète, que rien n’en transpira d’abord. Cependant, Virginie, qui avait encouragé sa soeur par mille douces représentations, poussa la complaisance jusqu’à se glisser auprès de la porte de la chambre à coucher de sa mère, chez laquelle la discussion avait lieu, pour y recueillir quelques phrases. Au premier voyage qu’elle fit du troisième au second étage, elle entendit son père qui s’écriait :\\n\\n— Madame, vous voulez donc tuer votre fille ?\\n\\n— Ma pauvre enfant, dit Virginie à sa soeur éplorée, papa prend ta défense !\\n\\n— Et que veulent-ils faire à Théodore ? demanda l’innocente créature.\\n\\nLa curieuse Virginie redescendit alors ; mais cette fois elle resta plus long-temps : elle apprit que Lebas aimait Augustine. Il était écrit que, dans cette mémorable journée, une maison ordinairement si calme serait un enfer. Monsieur Guillaume désespéra Joseph Lebas en lui confiant l’amour d’Augustine pour un étranger. Lebas, qui avait averti son ami de demander mademoiselle Virginie en mariage, vit ses espérances renversées. Mademoiselle Virginie, accablée de savoir que Joseph l’avait en quelque sorte refusée, fut prise d’une migraine. La zizanie, semée entre les deux époux par l’explication que monsieur et madame Guillaume avaient eue ensemble, et où, pour la troisième fois de leur vie, ils se trouvèrent d’opinions différentes, se manifesta d’une manière terrible. Enfin, à quatre heures après midi, Augustine, pâle, tremblante et les yeux rouges, comparut devant son père et sa mère. La pauvre enfant raconta naïvement la trop courte histoire de ses amours. Rassurée par l’allocution de son père, qui lui avait promis de l’écouter en silence, elle prit un certain courage en prononçant devant ses parents le nom de son cher Théodore de Sommervieux, et en fit malicieusement sonner la particule aristocratique. En se livrant au charme inconnu de parler de ses sentiments, elle trouva assez de hardiesse pour déclarer avec une innocente fermeté qu’elle aimait monsieur de Sommervieux, qu’elle le lui avait écrit, et ajouta, les larmes aux yeux :\\n\\n— Ce serait faire mon malheur que de me sacrifier à un autre.\\n\\n— Mais, Augustine, vous ne savez donc pas ce que c’est qu’un peintre ? s’écria sa mère avec horreur.\\n\\n— Madame Guillaume ! dit le vieux père en imposant silence à sa femme.\\n\\n— Augustine, dit-il, les artistes sont en général des meure-de-faim. Ils sont trop dépensiers pour ne pas être toujours de mauvais sujets. J’ai fourni feu M. Joseph Vernet, feu M. Lekain et feu M. Noverre. Ah ! si tu savais combien ce M. Noverre, M. le chevalier de Saint-Georges, et surtout M. Philidor, ont joué de tours à ce pauvre père Chevrel ! Ce sont de drôles de corps, je le sais bien. Ça vous a tous un babil, des manières… Ah ! jamais ton monsieur Sumer… Somm…\\n\\n— De Sommervieux, mon père !\\n\\n— Eh bien ! de Sommervieux, soit ! Jamais il n’aura été aussi agréable avec toi que M. le chevalier de Saint-Georges le fut avec moi, le jour où j’obtins une sentence des consuls contre lui. Aussi était-ce des gens de qualité d’autrefois.\\n\\n— Mais, mon père, monsieur Théodore est noble, et m’a écrit qu’il était riche. Son père s’appelait le chevalier de Sommervieux avant la révolution.\\n\\nA ces paroles, monsieur Guillaume regarda sa terrible moitié, qui, en femme contrariée frappait le plancher du bout du pied et gardait un morne silence. Elle évitait même de jeter ses yeux courroucés sur Augustine, et semblait laisser à monsieur Guillaume toute la responsabilité d’une affaire si grave, puisque ses avis n’étaient pas écoutés. Cependant, malgré son flegme apparent, quand elle vit son mari prenant si doucement son parti sur une catastrophe qui n’avait rien de commercial, elle s’écria :\\n\\n\\n\\n\\n\\n— En vérité, monsieur, vous êtes d’une faiblesse avec vos filles… mais…\\n\\nLe bruit d’une voiture qui s’arrêtait à la porte interrompit tout à coup la mercuriale que le vieux négociant redoutait déjà . En un moment, madame Roguin se trouva au milieu de la chambre, et, regardant les trois acteurs de cette scène domestique :\\n\\n— Je sais tout, ma cousine, dit-elle d’un air de protection.\\n\\nMadame Roguin avait un défaut, celui de croire que la femme d’un notaire de Paris pouvait jouer le rôle d’une petite maîtresse.\\n\\n— Je sais tout, répéta-t-elle, et je viens dans l’arche de Noé, comme la colombe, avec la branche d’olivier. J’ai lu cette allégorie dans le Génie du christianisme, dit-elle en se retournant vers madame Guillaume, la comparaison doit vous plaire, ma cousine. Savez-vous, ajouta-t-elle en souriant à Augustine, que ce monsieur de Sommervieux est un homme charmant ? Il m’a donné ce matin mon portrait fait de main de maître. Cela vaut au moins six mille francs.\\n\\nA ces mots, elle frappa doucement sur les bras de monsieur Guillaume. Le vieux négociant ne put s’empêcher de faire avec ses lèvres une grosse moue qui lui était particulière.\\n\\n— Je connais beaucoup monsieur de Sommervieux, reprit la colombe. Depuis une quinzaine de jours il vient à mes soirées, il en fait le charme. Il m’a conté toutes ses peines et m’a prise pour avocat. Je sais de ce matin qu’il adore Augustine, et il l’aura. Ah ! cousine, n’agitez pas ainsi la tête en signe de refus. Apprenez qu’il sera créé baron, et qu’il vient d’être nommé chevalier de la Légion-d’Honneur par l’empereur lui-même, au Salon. Roguin est devenu son notaire et connaît ses affaires. Eh bien ! monsieur de Sommervieux possède en bons biens au soleil douze mille livres de rente. Savez-vous que le beau-père d’un homme comme lui peut devenir quelque chose, maire de son arrondissement, par exemple ! N’avez-vous pas vu monsieur Dupont être fait comte de l’empire et sénateur pour être venu, en sa qualité de maire, complimenter l’empereur sur son entrée à Vienne. Oh ! ce mariage-là se fera. Je l’adore, moi, ce bon jeune homme. Sa conduite envers Augustine ne se voit que dans les romans. Va, ma petite, tu seras heureuse, et tout le monde voudrait être à ta place. J’ai chez moi, à mes soirées, madame la duchesse de Carigliano qui raffole de monsieur de Sommervieux. Quelques méchantes langues disent qu’elle ne vient chez moi que pour lui, comme si une duchesse d’hier était déplacée chez une Chevrel dont la famille a cent ans de bonne bourgeoisie.\\n\\n— Augustine, reprit madame Roguin après une petite pause, j’ai vu le portrait. Dieu ! qu’il est beau. Sais-tu que l’empereur a voulu le voir ? Il a dit en riant au Vice-Connétable que s’il y avait beaucoup de femmes comme celle-là à sa cour pendant qu’il y venait tant de rois, il se faisait fort de maintenir toujours la paix en Europe. Est-ce flatteur ?\\n\\nLes orages par lesquels cette journée avait commencé devaient ressembler à ceux de la nature, en ramenant un temps calme et serein. Madame Roguin déploya tant de séductions dans ses discours, elle sut attaquer tant de cordes à la fois dans les coeurs secs de monsieur et de madame Guillaume, qu’elle finit par en trouver une dont elle tira parti. A cette singulière époque, le commerce et la finance avaient plus que jamais la folle manie de s’allier aux grands seigneurs, et les généraux de l’empire profitèrent assez bien de ces dispositions. Monsieur Guillaume s’élevait singulièrement contre cette déplorable passion. Ses axiomes favoris étaient que, pour trouver le bonheur, une femme devait épouser un homme de sa classe ; on était toujours tôt ou tard puni d’avoir voulu monter trop haut ; l’amour résistait si peu aux tracas du ménage, qu’il fallait trouver l’un chez l’autre des qualités bien solides pour être heureux ; il ne fallait pas que l’un des deux époux en sût plus que l’autre, parce qu’on devait avant tout se comprendre ; un mari qui parlait grec et la femme latin, risquaient de mourir de faim. Il avait inventé cette espèce de proverbe. Il comparait les mariages ainsi faits à ces anciennes étoffes de soie et de laine, dont la soie finissait toujours par couper la laine. Cependant, il se trouve tant de vanité au fond du coeur de l’homme, que la prudence du pilote qui gouvernait si bien le Chat-qui-pelote, succomba sous l’agressive volubilité de madame Roguin. La sévère madame Guillaume, la première, trouva dans l’inclination de sa fille des motifs pour déroger à ces principes, et pour consentir à recevoir au logis monsieur de Sommervieux, qu’elle se promit de soumettre à un rigoureux examen.\\n\\nLe vieux négociant alla trouver Joseph Lebas, et l’instruisit de l’état des choses. A six heures et demie, la salle à manger illustrée par le peintre, réunit sous son toit de verre, madame et monsieur Roguin, le jeune peintre et sa charmante Augustine, Joseph Lebas qui prenait son bonheur en patience, et mademoiselle Virginie dont la migraine avait cessé. Monsieur et madame Guillaume virent en perspective leurs enfants établis et les destinées du Chat-qui-pelote remises en des mains habiles. Leur contentement fut au comble, quand, au dessert, Théodore leur fit présent de l’étonnant tableau qu’ils n’avaient pu voir, et qui représentait l’intérieur de cette vieille boutique, à laquelle était dû tant de bonheur.\\n\\n— C’est-y gentil, s’écria Guillaume. Dire qu’on voulait donner trente mille francs de cela.\\n\\n— Mais c’est qu’on y trouve mes barbes, reprit madame Guillaume.\\n\\n— Et ces étoffes dépliées, ajouta Lebas, on les prendrait avec la main.\\n\\n— Les draperies font toujours très-bien, répondit le peintre. Nous serions trop heureux, nous autres artistes modernes, d’atteindre à la perfection de la draperie antique.\\n\\n— Vous aimez donc la draperie, s’écria le père Guillaume. Eh bien, sarpejeu ! touchez là , mon jeune ami. Puisque vous estimez le commerce, nous nous entendrons. Eh ! pourquoi le mépriserait-on ? Le monde a commencé par là , puisque Adam a vendu le paradis pour une pomme. Ça n’a pas été une fameuse spéculation, par exemple !\\n\\nEt le vieux négociant se mit à éclater d’un gros rire franc excité par le vin de Champagne qu’il faisait circuler généreusement. Le bandeau qui couvrait les yeux du jeune artiste fut si épais qu’il trouva ses futurs parents aimables. Il ne dédaigna pas de les égayer par quelques charges de bon goût. Aussi plut-il généralement. Le soir, quand le salon meublé de choses très-cossues, pour se servir de l’expression de Guillaume, fut désert ; pendant que madame Guillaume s’en allait de table en cheminée, de candélabre en flambeau, soufflant avec précipitation les bougies, le brave négociant, qui savait toujours voir clair aussitôt qu’il s’agissait d’affaires ou d’argent, attira sa fille Augustine auprès de lui ; puis, après l’avoir prise sur ses genoux, il lui tint ce discours :\\n\\n— Ma chère enfant, tu épouseras ton Sommervieux, puisque tu le veux ; permis à toi de risquer ton capital de bonheur. Mais je ne me laisse pas prendre à ces trente mille francs que l’on gagne à gâter de bonnes toiles. L’argent qui vient si vite s’en va de même. N’ai-je pas entendu dire ce soir à ce jeune écervelé que si l’argent était rond, c’était pour rouler ! S’il est rond pour les gens prodigues, il est plat pour les gens économes qui l’empilent et l’amassent. Or, mon enfant, ce beau garçon-là parle de te donner des voitures, des diamants ? Il a de l’argent, qu’il le dépense pour toi ! bene sit ! Je n’ai rien à y voir. Mais quant à ce que je te donne, je ne veux pas que des écus si péniblement ensachés s’en aillent en carrosses ou en colifichets. Qui dépense trop n’est jamais riche. Avec les cent mille écus de sa dot on n’achète pas encore tout Paris. Tu as beau avoir à recueillir un jour quelques centaines de mille francs, je te les ferai attendre, sarpejeu ! le plus long-temps possible. J’ai donc attiré ton prétendu dans un coin, et un homme qui a mené la faillite Lecocq n’a pas eu grande peine à faire consentir un artiste à se marier séparé de biens avec sa femme. J’aurai l’oeil au contrat pour bien faire stipuler les donations qu’il se propose de te constituer. Allons, mon enfant, j’espère être grand-père, sarpejeu ! je veux m’occuper déjà de mes petits-enfants : jure-moi donc ici de ne jamais rien signer en fait d’argent que par mon conseil ; et si j’allais trouver trop tôt le père Chevrel, jure-moi de consulter le jeune Lebas, ton beau-frère. Promets-le-moi.\\n\\n— Oui, mon père, je vous le jure.\\n\\nA ces mots prononcés d’une voix douce, le vieillard baisa sa fille sur les deux joues. Ce soir-là , tous les amants dormirent presque aussi paisiblement que monsieur et madame Guillaume. Quelques mois après ce mémorable dimanche, le maître-autel de Saint-Leu fut témoin de deux mariages bien différents. Augustine et Théodore s’y présentèrent dans tout l’éclat du bonheur, les yeux pleins d’amour, parés de toilettes élégantes, attendus par un brillant équipage. Venue dans un bon remise avec sa famille, Virginie, donnant le bras à son père, suivait sa jeune soeur humblement et dans de plus simples atours, comme une ombre nécessaire aux harmonies de ce tableau. Monsieur Guillaume s’était donné toutes les peines imaginables pour obtenir à l’église que Virginie fût mariée avant Augustine ; mais il eut la douleur de voir le haut et le bas clergé s’adresser en toute circonstance à la plus élégante des mariées. Il entendit quelques-uns de ses voisins approuver singulièrement le bon sens de mademoiselle Virginie, qui faisait, disaient-ils, le mariage le plus solide, et restait fidèle au quartier ; tandis qu’ils lancèrent quelques brocards suggérés par l’envie sur Augustine qui épousait un artiste, un noble ; ils ajoutèrent avec une sorte d’effroi que, si les Guillaume avaient de l’ambition, la draperie était perdue. Un vieux marchand d’éventails ayant dit que ce mange-tout-là l’aurait bientôt mise sur la paille, le père Guillaume s’applaudit in petto de la prudence qu’il avait mise dans la rédaction des conventions matrimoniales. Le soir, la famille se sépara après un bal somptueux, suivi d’un de ces soupers plantureux dont le souvenir commence à se perdre dans la génération présente. Monsieur et madame Guillaume restèrent dans leur hôtel de la rue du Colombier où la noce avait eu lieu. Monsieur et madame Lebas retournèrent dans leur remise à la vieille maison de la rue Saint-Denis pour y diriger la nauf du Chat-qui-pelote. L’artiste, ivre de bonheur, prit entre ses bras sa chère Augustine, l’enleva vivement quand leur coupé arriva rue des Trois-Frères, et la porta dans son élégant appartement.\\n\\nLa fougue de passion qui possédait Théodore fit dévorer au jeune ménage près d’une année entière sans que le moindre nuage vînt altérer l’azur du ciel sous lequel ils vivaient. Pour eux, l’existence n’eut rien de pesant. Théodore répandait sur chaque journée d’incroyables fioriture de plaisirs. Il se plaisait à varier les emportements de la passion, par la molle langueur de ces repos où les âmes sont lancées si haut dans l’extase qu’elles semblent y oublier l’union corporelle. Incapable de réfléchir, l’heureuse Augustine se prêtait à l’allure onduleuse de son bonheur. Elle ne croyait pas faire encore assez en se livrant toute à l’amour permis et saint du mariage. Simple et naïve, elle ne connaissait ni la coquetterie des refus, ni l’empire qu’une jeune demoiselle du grand monde se crée sur un mari par d’adroits caprices. Elle aimait trop pour calculer l’avenir, et n’imaginait pas qu’une vie si délicieuse pût jamais cesser. Heureuse d’être alors tous les plaisirs de son mari, elle crut que cet inextinguible amour serait toujours pour elle la plus belle de toutes les parures, comme son dévouement et son obéissance seraient un éternel attrait. Enfin, la félicité de l’amour l’avait rendue si brillante, que sa beauté lui inspira de l’orgueil et lui donna la conscience de pouvoir toujours régner sur un homme aussi facile à enflammer que monsieur de Sommervieux. Ainsi son état de femme ne lui apporta d’autres enseignements que ceux de l’amour. Au sein de ce bonheur, elle resta l’ignorante petite fille qui vivait obscurément rue Saint-Denis, et ne pensa point à prendre les manières, l’instruction, le ton du monde dans lequel elle devait vivre. Ses paroles étant des paroles d’amour, elle y déployait bien une sorte de souplesse d’esprit et une certaine délicatesse d’expression ; mais elle se servait du langage commun à toutes les femmes quand elles se trouvent plongées dans une passion qui semble être leur élément. Si, par hasard, une idée discordante avec celles de Théodore était exprimée par Augustine, le jeune artiste en riait comme on rit des premières fautes que fait un étranger, mais qui finissent par fatiguer s’il ne se corrige pas.\\n\\nCependant, à l’expiration de cette année aussi charmante que rapide, Sommervieux sentit un matin la nécessité de reprendre ses travaux et ses habitudes. Sa femme était enceinte. Il revit ses amis. Pendant les longues souffrances de l’année où, pour la première fois, une jeune femme nourrit un enfant, il travailla sans doute avec ardeur ; mais parfois il retourna chercher quelques distractions dans le grand monde. La maison où il allait le plus volontiers était celle de la duchesse de Carigliano qui avait fini par attirer chez elle le célèbre artiste. Quand Augustine fut rétablie, quand son fils ne réclama plus ces soins assidus qui interdisent à une mère les plaisirs du monde, Théodore en était arrivé à vouloir éprouver cette jouissance d’amour-propre que nous donne la société quand nous y apparaissons avec une belle femme, objet d’envie et d’admiration. Parcourir les salons en s’y montrant avec l’éclat emprunté de la gloire de son mari, se voir jalousée par toutes les femmes, fut pour Augustine une nouvelle moisson de plaisirs ; mais ce fut le dernier reflet que devait jeter son bonheur conjugal. Elle commença par offenser la vanité de son mari, quand, malgré de vains efforts, elle laissa percer son ignorance, l’impropriété de son langage et l’étroitesse de ses idées. Le caractère de Sommervieux, dompté pendant près de deux ans et demi par les premiers emportements de l’amour, reprit, avec la tranquillité d’une possession moins jeune, sa pente et ses habitudes un moment détournées de leur cours. La poésie, la peinture et les exquises jouissances de l’imagination possèdent sur les esprits élevés des droits imprescriptibles. Ces besoins d’une âme forte n’avaient pas été trompés chez Théodore pendant ces deux années, ils avaient trouvé seulement une pâture nouvelle. Quand les champs de l’amour furent parcourus, quand l’artiste eut, comme les enfants, cueilli des roses et des bleuets avec une telle avidité qu’il ne s’apercevait pas que ses mains ne pouvaient plus les tenir, la scène changea. Si le peintre montrait à sa femme les croquis de ses plus belles compositions, il l’entendait s’écrier comme eût fait le père Guillaume : « C’est bien joli ! » Son admiration sans chaleur ne provenait pas d’un sentiment consciencieux, mais de la croyance sur parole de l’amour. Augustine préférait un regard au plus beau tableau. Le seul sublime qu’elle connût était celui du coeur. Enfin, Théodore ne put se refuser à l’évidence d’une vérité cruelle : sa femme n’était pas sensible à la poésie, elle n’habitait pas sa sphère, elle ne le suivait pas dans tous ses caprices, dans ses improvisations, dans ses joies, dans ses douleurs ; elle marchait terre à terre dans le monde réel, tandis qu’il avait la tête dans les cieux. Les esprits ordinaires ne peuvent pas apprécier les souffrances renaissantes de l’être qui, uni à un autre par le plus intime de tous les sentiments, est obligé de refouler sans cesse les plus chères expansions de sa pensée, et de faire rentrer dans le néant les images qu’une puissance magique le force à créer. Pour lui, ce supplice est d’autant plus cruel, que le sentiment qu’il porte à son compagnon ordonne, par sa première loi, de ne jamais rien se dérober l’un à l’autre, et de confondre les effusions de la pensée aussi bien que les épanchements de l’âme. On ne trompe pas impunément les volontés de la nature : elle est inexorable comme la Nécessité, qui, certes, est une sorte de nature sociale. Sommervieux se réfugia dans le calme et le silence de son atelier, en espérant que l’habitude de vivre avec des artistes pourrait former sa femme, et développerait en elle les germes de haute intelligence engourdis que quelques esprits supérieurs croient préexistants chez tous les êtres ; mais Augustine était trop sincèrement religieuse pour ne pas être effrayée du ton des artistes. Au premier dîner que donna Théodore, elle entendit un jeune peintre disant avec cette enfantine légèreté qu’elle ne sut pas reconnaître et qui absout une plaisanterie de toute irréligion : — Mais, madame, votre paradis n’est pas plus beau que la Transfiguration de Raphaël ? Eh ! bien, je me suis lassé de la regarder. Augustine apporta donc dans cette société spirituelle un esprit de défiance qui n’échappait à personne. Elle gêna. Les artistes gênés sont impitoyables : ils fuient ou se moquent. Madame Guillaume avait, entre autres ridicules, celui d’outrer la dignité qui lui semblait l’apanage d’une femme mariée ; et quoiqu’elle s’en fût souvent moquée, Augustine ne sut pas se défendre d’une légère imitation de la pruderie maternelle. Cette exagération de pudeur, que n’évitent pas toujours les femmes vertueuses, suggéra quelques épigrammes à coups de crayon dont l’innocent badinage était de trop bon goût pour que Sommervieux pût s’en fâcher. Ces plaisanteries eussent été même plus cruelles, elles n’étaient après tout que des représailles exercées sur lui par ses amis. Mais rien ne pouvait être léger pour une âme qui recevait aussi facilement que celle de Théodore des impressions étrangères. Aussi éprouva-t-il insensiblement une froideur qui ne pouvait aller qu’en croissant. Pour arriver au bonheur conjugal, il faut gravir une montagne dont l’étroit plateau est bien près d’un revers aussi rapide que glissant, et l’amour du peintre le descendait. Il jugea sa femme incapable d’apprécier les considérations morales qui justifiaient, à ses propres yeux, la singularité de ses manières envers elle, et se crut fort innocent en lui cachant des pensées qu’elle ne comprenait pas et des écarts peu justifiables au tribunal d’une conscience bourgeoise. Augustine se renferma dans une douleur morne et silencieuse. Ces sentiments secrets mirent entre les deux époux un voile qui devait s’épaissir de jour en jour. Sans que son mari manquât d’égards envers elle, Augustine ne pouvait s’empêcher de trembler en le voyant réserver pour le monde les trésors d’esprit et de grâce qu’il venait jadis mettre à ses pieds. Bientôt, elle interpréta fatalement les discours spirituels qui se tiennent dans le monde sur l’inconstance des hommes. Elle ne se plaignit pas, mais son attitude équivalait à des reproches. Trois ans après son mariage, cette femme jeune et jolie qui passait si brillante dans son brillant équipage, qui vivait dans une sphère de gloire et de richesse enviée de tant de gens insouciants et incapables d’apprécier justement les situations de la vie, fut en proie à de violents chagrins. Ses couleurs pâlirent. Elle réfléchit, elle compara ; puis, le malheur lui déroula les premiers textes de l’expérience. Elle résolut de rester courageusement dans le cercle de ses devoirs, en espérant que cette conduite généreuse lui ferait recouvrer tôt ou tard l’amour de son mari ; mais il n’en fut pas ainsi. Quand Sommervieux, fatigué de travail, sortait de son atelier, Augustine ne cachait pas si promptement son ouvrage, que le peintre ne pût apercevoir sa femme raccommodant avec toute la minutie d’une bonne ménagère le linge de la maison et le sien. Elle fournissait, avec générosité, sans murmure, l’argent nécessaire aux prodigalités de son mari ; mais, dans le désir de conserver la fortune de son cher Théodore, elle se montrait économe soit pour elle, soit dans certains détails de l’administration domestique. Cette conduite est incompatible avec le laisser-aller des artistes qui, sur la fin de leur carrière, ont tant joui de la vie, qu’ils ne se demandent jamais la raison de leur ruine. Il est inutile de marquer chacune des dégradations de couleur par lesquelles la teinte brillante de leur lune de miel atteignit à une profonde obscurité. Un soir, la triste Augustine, qui depuis long-temps entendait son mari parler avec enthousiasme de madame la duchesse de Carigliano, reçut d’une amie quelques avis méchamment charitables sur la nature de l’attachement qu’avait conçu Sommervieux pour cette célèbre coquette qui donnait le ton à la cour impériale. A vingt et un ans, dans tout l’éclat de la jeunesse et de la beauté, Augustine se vit trahie pour une femme de trente-six ans. En se sentant malheureuse au milieu du monde et de ses fêtes désertes pour elle, la pauvre petite ne comprit plus rien à l’admiration qu’elle y excitait, ni à l’envie qu’elle inspirait. Sa figure prit une nouvelle expression. La mélancolie versa dans ses traits la douceur de la résignation et la pâleur d’un amour dédaigné. Elle ne tarda pas à être courtisée par les hommes les plus séduisants ; mais elle resta solitaire et vertueuse. Quelques paroles de dédain, échappées à son mari, lui donnèrent un incroyable désespoir. Une lueur fatale lui fit entrevoir les défauts de contact qui, par suite des mesquineries de son éducation, empêchaient l’union complète de son âme avec celle de Théodore : elle eut assez d’amour pour l’absoudre et pour se condamner. Elle pleura des larmes de sang, et reconnut trop tard qu’il est des mésalliances d’esprit aussi bien que des mésalliances de moeurs et de rang. En songeant aux délices printanières de son union, elle comprit l’étendue du bonheur passé, et convint en elle même qu’une si riche moisson d’amour était une vie entière qui ne pouvait se payer que par du malheur. Cependant elle aimait trop sincèrement pour perdre toute espérance. Aussi osa-t-elle entreprendre à vingt et un ans de s’instruire et de rendre son imagination au moins digne de celle qu’elle admirait.\\n\\n— Si je ne suis pas poète, se disait-elle, au moins je comprendrai la poésie.\\n\\nEt déployant alors cette force de volonté, cette énergie que les femmes possèdent toutes quand elles aiment, madame de Sommervieux tenta de changer son caractère, ses moeurs et ses habitudes ; mais en dévorant des volumes, en apprenant avec courage, elle ne réussit qu’à devenir moins ignorante. La légèreté de l’esprit et les grâces de la conversation sont un don de la nature ou le fruit d’une éducation commencée au berceau. Elle pouvait apprécier la musique, en jouir, mais non chanter avec goût. Elle comprit la littérature et les beautés de la poésie, mais il était trop tard pour en orner sa rebelle mémoire. Elle entendait avec plaisir les entretiens du monde, mais elle n’y fournissait rien de brillant. Ses idées religieuses et ses préjugés d’enfance s’opposèrent à la complète émancipation de son intelligence. Enfin, il s’était glissé contre elle, dans l’âme de Théodore, une prévention qu’elle ne put vaincre. L’artiste se moquait de ceux qui lui vantaient sa femme, et ses plaisanteries étaient assez fondées : il imposait tellement à cette jeune et touchante créature, qu’en sa présence, ou en tête-à -tête, elle tremblait. Embarrassée par son trop grand désir de plaire, elle sentait son esprit et ses connaissances s’évanouir dans un seul sentiment. La fidélité d’Augustine déplut même à cet infidèle mari, qui semblait l’engager à commettre des fautes en taxant sa vertu d’insensibilité. Augustine s’efforça en vain d’abdiquer sa raison, de se plier aux caprices, aux fantaisies de son mari, et de se vouer à l’égoïsme de sa vanité ; elle ne recueillit point le fruit de ces sacrifices. Peut-être avaient-ils tous deux laissé passer le moment où les âmes peuvent se comprendre. Un jour le coeur trop sensible de la jeune épouse reçut un de ces coups qui font si fortement plier les liens du sentiment, qu’on peut les croire rompus. Elle s’isola. Mais bientôt une fatale pensée lui suggéra d’aller chercher des consolations et des conseils au sein de sa famille.\\n\\nUn matin donc, elle se dirigea vers la grotesque façade de l’humble et silencieuse maison où s’était écoulée son enfance. Elle soupira en revoyant cette croisée d’où, un jour, elle avait envoyé un premier baiser à celui qui répandait aujourd’hui sur sa vie autant de gloire que de malheur. Rien n’était changé dans l’antre où se rajeunissait cependant le commerce de la draperie. La soeur d’Augustine occupait au comptoir antique la place de sa mère. La jeune affligée rencontra son beau-frère la plume derrière l’oreille. Elle fut à peine écoutée, tant il avait l’air affairé. Les redoutables signaux d’un inventaire général se faisaient autour de lui. Aussi la quitta-t-il en la priant d’excuser. Elle fut reçue assez froidement par sa soeur, qui lui manifesta quelque rancune. En effet, Augustine, brillante et descendant d’un joli équipage, n’était jamais venue voir sa soeur qu’en passant. La femme du prudent Lebas s’imagina que l’argent était la cause première de cette visite matinale, elle essaya de se maintenir sur un ton de réserve qui fit sourire plus d’une fois Augustine. La femme du peintre vit que, sauf les barbes au bonnet, sa mère avait trouvé dans Virginie un successeur qui conservait l’antique honneur du Chat-qui-pelote. Au déjeuner, elle aperçut, dans le régime de la maison, certains changements qui faisaient honneur au bon sens de Joseph Lebas : les commis ne se levèrent pas au dessert, on leur laissait la faculté de parler, et l’abondance de la table annonçait une aisance sans luxe. La jeune élégante trouva les coupons d’une loge aux Français où elle se souvint d’avoir vu sa soeur de loin en loin. Madame Lebas avait sur les épaules un cachemire dont la magnificence attestait la générosité avec laquelle son mari s’occupait d’elle. Enfin, les deux époux marchaient avec leur siècle. Augustine fut bientôt pénétrée d’attendrissement, en reconnaissant, pendant les deux tiers de cette journée, le bonheur égal, sans exaltation, il est vrai, mais aussi sans orages, que goûtait ce couple convenablement assorti. Ils avaient accepté la vie comme une entreprise commerciale où il s’agissait de faire, avant tout, honneur à ses affaires. La femme, n’ayant pas rencontré dans son mari un amour excessif, s’était appliquée à le faire naître. Insensiblement amené à estimer, à chérir Virginie, le temps que le bonheur mit à éclore, fut, pour Joseph Lebas et pour sa femme, un gage de durée. Aussi, lorsque la plaintive Augustine exposa sa situation douloureuse, eut-elle à essuyer le déluge de lieux communs que la morale de la rue Saint-Denis fournissait à sa soeur.\\n\\n— Le mal est fait, ma femme, dit Joseph Lebas, il faut chercher à donner de bons conseils à notre soeur. Puis, l’habile négociant analysa lourdement les ressources que les lois et les moeurs pouvaient offrir à Augustine pour sortir de cette crise ; il en numérota pour ainsi dire les considérations, les rangea par leur force dans des espèces de catégories, comme s’il se fût agi de marchandises de diverses qualités ; puis il les mit en balance, les pesa, et conclut en développant la nécessité où était sa belle-soeur de prendre un parti violent qui ne satisfit point l’amour qu’elle ressentait encore pour son mari. Aussi ce sentiment se réveilla-t-il dans toute sa force quand elle entendit Joseph Lebas parlant de voies judiciaires. Elle remercia ses deux amis, et revint chez elle encore plus indécise qu’elle ne l’était avant de les avoir consultés. Elle hasarda de se rendre alors à l’antique hôtel de la rue du Colombier, dans le dessein de confier ses malheurs à son père et à sa mère. La pauvre petite femme ressemblait à ces malades qui, arrivés à un état désespéré, essaient de toutes les recettes et se confient même aux remèdes de bonne femme. Les deux vieillards la reçurent avec une effusion de sentiment qui l’attendrit. Cette visite leur apportait une distraction qui, pour eux, valait un trésor. Depuis quatre ans, ils marchaient dans la vie comme des navigateurs sans but et sans boussole. Assis au coin de leur feu, ils se racontaient l’un à l’autre tous les désastres du Maximum, leurs anciennes acquisitions de draps, la manière dont ils avaient évité les banqueroutes, et surtout cette célèbre faillite Lecocq, la bataille de Marengo du père Guillaume. Puis, quand ils avaient épuisé les vieux procès, ils récapitulaient les additions de leurs inventaires les plus productifs, et se narraient encore les vieilles histoires du quartier Saint-Denis. A deux heures, le père Guillaume allait donner un coup d’oeil à l’établissement du Chat-qui-pelote. En revenant il s’arrêtait à toutes les boutiques, autrefois ses rivales, et dont les jeunes propriétaires espéraient entraîner le vieux négociant dans quelque escompte aventureux, que, selon sa coutume, il ne refusait jamais positivement. Deux bons chevaux normands mouraient de gras-fondu dans l’écurie de l’hôtel ; madame Guillaume ne s’en servait que pour se faire traîner tous les dimanches à la grand’messe de sa paroisse. Trois fois par semaine ce respectable couple tenait table ouverte. Grâce à l’influence de son gendre Sommervieux, le père Guillaume avait été nommé membre du comité consultatif pour l’habillement des troupes. Depuis que son mari s’était ainsi trouvé placé haut dans l’administration, madame Guillaume avait pris la détermination de représenter. Leurs appartements étaient encombrés de tant d’ornements d’or et d’argent, et de meubles sans goût mais de valeur certaine, que la pièce la plus simple y ressemblait à une chapelle. L’économie et la prodigalité semblaient se disputer dans chacun des accessoires de cet hôtel. L’on eût dit que monsieur Guillaume avait eu en vue de faire un placement d’argent jusque dans l’acquisition d’un flambeau. Au milieu de ce bazar, dont la richesse accusait le désoeuvrement des deux époux, le célèbre tableau de Sommervieux avait obtenu la place d’honneur. Il faisait la consolation de monsieur et de madame Guillaume qui tournaient vingt fois par jour leurs yeux harnachés de bésicles vers cette image de leur ancienne existence, pour eux si active et si amusante. L’aspect de cet hôtel et de ces appartements où tout avait une senteur de vieillesse et de médiocrité, le spectacle donné par ces deux êtres qui semblaient échoués sur un rocher d’or loin du monde et des idées qui font vivre, surprirent Augustine. Elle contemplait en ce moment la seconde partie du tableau dont le commencement l’avait frappée chez Joseph Lebas, celui d’une vie agitée quoique sans mouvement, espèce d’existence mécanique et instinctive semblable à celle des castors. Elle eut alors je ne sais quel orgueil de ses chagrins, en pensant qu’ils prenaient leur source dans un bonheur de dix-huit mois qui valait à ses yeux mille existences comme celle dont le vide lui semblait horrible. Cependant elle cacha ce sentiment peu charitable, et déploya pour ses vieux parents les grâces nouvelles de son esprit, les coquetteries de tendresse que l’amour lui avait révélées, et les disposa favorablement à écouter ses doléances matrimoniales. Les vieilles gens ont un faible pour ces sortes de confidences. Madame Guillaume voulut être instruite des plus légers détails de cette vie étrange qui, pour elle, avait quelque chose de fabuleux. Les voyages du baron de La Hontan, qu’elle commençait toujours sans jamais les achever, ne lui apprirent rien de plus inouï sur les sauvages du Canada.\\n\\n— Comment, mon enfant, ton mari s’enferme avec des femmes nues, et tu as la simplicité de croire qu’il les dessine ?\\n\\nA cette exclamation, la grand’mère posa ses lunettes sur une petite travailleuse, secoua ses jupons et plaça ses mains jointes sur ses genoux élevés par une chaufferette, son piédestal favori.\\n\\n— Mais, ma mère, tous les peintres sont obligés d’avoir des modèles.\\n\\n— Il s’est bien gardé de nous dire tout cela quand il t’a demandée en mariage. Si je l’avais su, je n’aurais pas donné ma fille à un homme qui fait un pareil métier. La religion défend ces horreurs-là , ça n’est pas moral. A quelle heure nous disais-tu donc qu’il rentre chez lui ?\\n\\n— Mais à une heure, deux heures…\\n\\nLes deux époux se regardèrent dans un profond étonnement.\\n\\n— Il joue donc ? dit monsieur Guillaume. Il n’y avait que les joueurs qui, de mon temps, rentrassent si tard.\\n\\nAugustine fit une petite moue qui repoussait cette accusation.\\n\\n— Il doit te faire passer de cruelles nuits à l’attendre, reprit madame Guillaume. Mais, non, tu te couches, n’est-ce pas ? Et quand il a perdu, le monstre te réveille.\\n\\n— Non, ma mère, il est au contraire quelquefois très-gai. Assez souvent même, quand il fait beau, il me propose de me lever pour aller dans les bois.\\n\\n— Dans les bois, à ces heures-là ? Tu as donc un bien petit appartement qu’il n’a pas assez de sa chambre, de ses salons, et qu’il lui faille ainsi courir pour… Mais c’est pour t’enrhumer, que le scélérat te propose ces parties-là . Il veut se débarrasser de toi. A-t-on jamais vu un homme établi, qui a un commerce tranquille, galoper comme un loup-garou ?\\n\\n— Mais, ma mère, vous ne comprenez donc pas que, pour développer son talent, il a besoin d’exaltation. Il aime beaucoup les scènes qui…\\n\\n— Ah ! je lui en ferais de belles, des scènes, moi, s’écria madame Guillaume en interrompant sa fille. Comment peux-tu garder des ménagements avec un homme pareil ? D’abord, je n’aime pas qu’il ne boive que de l’eau. Ça n’est pas sain. Pourquoi montre-t-il de la répugnance à voir les femmes quand elles mangent ? Quel singulier genre ! Mais c’est un fou. Tout ce que tu nous en as dit n’est pas possible, Un homme ne peut pas partir de sa maison sans souffler mot et ne revenir que dix jours après. Il te dit qu’il a été à Dieppe pour peindre la mer. Est-ce qu’on peint la mer ? Il te fait des contes à dormir debout.\\n\\nAugustine ouvrit la bouche pour défendre son mari ; mais madame Guillaume lui imposa silence par un geste de main auquel un reste d’habitude la fit obéir, et sa mère s’écria d’un ton sec :\\n\\n— Tiens, ne me parle pas de cet homme-là ! il n’a jamais mis le pied dans une église que pour te voir et t’épouser. Les gens sans religion sont capables de tout. Est-ce que Guillaume s’est jamais avisé de me cacher quelque chose, de rester des trois jours sans me dire ouf, et de babiller ensuite comme une pie borgne ?\\n\\n— Ma chère mère, vous jugez trop sévèrement les gens supérieurs. S’ils avaient des idées semblables à celles des autres, ce ne seraient plus des gens à talent.\\n\\n— Eh bien ! que les gens à talent restent chez eux et ne se marient pas. Comment ! un homme à talent rendra sa femme malheureuse ! et parce qu’il a du talent, ce sera bien ? Talent, talent ! Il n’y a pas tant de talent à dire comme lui blanc et noir à toute minute, à couper la parole aux gens, à battre du tambour chez soi, à ne jamais vous laisser savoir sur quel pied danser, à forcer une femme de ne pas s’amuser avant que les idées de monsieur ne soient gaies, d’être triste, dès qu’il est triste.\\n\\n— Mais, ma mère, le propre de ces imaginations-là …\\n\\n— Qu’est-ce que c’est que ces imaginations-là ? reprit madame Guillaume en interrompant encore sa fille. Il en a de belles, ma foi ! Qu’est-ce qu’un homme auquel il prend tout à coup, sans consulter de médecin, la fantaisie de ne manger que des légumes ? Encore, si c’était par religion, sa diète lui servirait à quelque chose ; mais il n’en a pas plus qu’un huguenot. A-t-on jamais vu un homme aimer, comme lui, les chevaux plus qu’il n’aime son prochain, se faire friser les cheveux comme un païen, coucher des statues sous de la mousseline, faire fermer ses fenêtres le jour pour travailler à la lampe ? Tiens, laisse-moi, s’il n’était pas si grossièrement immoral, il serait bon à mettre aux Petites-Maisons. Consulte monsieur Loraux, le vicaire de Saint-Sulpice, demande-lui son avis sur tout cela, il te dira que ton mari ne se conduit pas comme un chrétien…\\n\\n— Oh ! ma mère ! pouvez-vous croire…\\n\\n— Oui, je le crois ! Tu l’as aimé, tu n’aperçois rien de ces choses-là . Mais, moi, vers les premiers temps de son mariage, je me souviens de l’avoir rencontré dans les Champs-Elysées. Il était à cheval. Eh bien ! il galopait par moment ventre à terre, et puis il s’arrêtait pour aller pas à pas. Je me suis dit alors : « Voilà un homme qui n’a pas de jugement. »\\n\\n— Ah ! s’écria monsieur Guillaume en se frottant les mains, comme j’ai bien fait de t’avoir mariée séparée de biens avec cet original-là !\\n\\nQuand Augustine eut l’imprudence de raconter les griefs véritables qu’elle avait à exposer contre son mari, les deux vieillards restèrent muets d’indignation. Le mot de divorce fut bientôt prononcé par madame Guillaume. Au mot de divorce, l’inactif négociant fut comme réveillé. Stimulé par l’amour qu’il avait pour sa fille, et aussi par l’agitation qu’un procès allait donner à sa vie sans événements, le père Guillaume prit la parole. Il se mit à la tête de la demande en divorce, la dirigea, plaida presque, il offrit à sa fille de se charger de tous les frais, de voir les juges, les avoués, les avocats, de remuer ciel et terre. Madame de Sommervieux, effrayée, refusa les services de son père, dit qu’elle ne voulait pas se séparer de son mari, dût-elle être dix fois plus malheureuse encore, et ne parla plus de ses chagrins. Après avoir été accablée par ses parents de tous ces petits soins muets et consolateurs par lesquels les deux vieillards essayèrent de la dédommager, mais en vain, de ses peines de coeur, Augustine se retira en sentant l’impossibilité de parvenir à faire bien juger les hommes supérieurs par des esprits faibles. Elle apprit qu’une femme devait cacher à tout le monde, même à ses parents, des malheurs pour lesquels on rencontre si difficilement des sympathies. Les orages et les souffrances des sphères élevées ne peuvent être appréciés que par les nobles esprits qui les habitent. En toute chose, nous ne pouvons être jugés que par nos pairs.\\n\\nLa pauvre Augustine se retrouva donc dans la froide atmosphère de son ménage, livrée à l’horreur de ses méditations. L’étude n’était plus rien pour elle, puisque l’étude ne lui avait pas rendu le coeur de son mari. Initiée aux secrets de ces âmes de feu mais privée de leurs ressources, elle participait avec force à leurs peines sans partager leurs plaisirs. Elle s’était dégoûtée du monde, qui lui semblait mesquin et petit devant les événements des passions. Enfin, sa vie était manquée. Un soir, elle fut frappée d’une pensée qui vint illuminer ses ténébreux chagrins comme un rayon céleste. Cette idée ne pouvait sourire qu’à un coeur aussi pur, aussi vertueux que l’était le sien. Elle résolut d’aller chez la duchesse de Carigliano, non pas pour lui redemander le coeur de son mari, mais pour s’y instruire des artifices qui le lui avaient enlevé ; mais pour intéresser à la mère des enfants de son ami cette orgueilleuse femme du monde ; mais pour la fléchir et la rendre complice de son bonheur à venir comme elle était l’instrument de son malheur présent.\\n\\nUn jour donc, la timide Augustine, armée d’un courage surnaturel, monta en voiture, à deux heures après midi, pour essayer de pénétrer jusqu’au boudoir de la célèbre coquette, qui n’était jamais visible avant cette heure-là . Madame de Sommervieux ne connaissait pas encore les antiques et somptueux hôtels du faubourg Saint-Germain. Quand elle parcourut ces vestibules majestueux, ces escaliers grandioses, ces salons immenses ornés de fleurs malgré les rigueurs de l’hiver, et décorés avec ce goût particulier aux femmes qui sont nées dans l’opulence ou avec les habitudes distinguées de l’aristocratie, Augustine eut un affreux serrement de coeur. Elle envia les secrets de cette élégance de laquelle elle n’avait jamais eu l’idée. Elle respira un air de grandeur qui lui expliqua l’attrait de cette maison pour son mari. Quand elle parvint aux petits appartements de la duchesse, elle éprouva de la jalousie et une sorte de désespoir, en y admirant la voluptueuse disposition des meubles, des draperies et des étoffes tendues. Là le désordre était une grâce, là le luxe affectait une espèce de dédain pour la richesse. Les parfums répandus dans cette douce atmosphère flattaient l’odorat sans l’offenser. Les accessoires de l’appartement s’harmoniaient avec une vue ménagée par des glaces sans tain sur les pelouses d’un jardin planté d’arbres verts. Tout était séduction, et le calcul ne s’y sentait point. Le génie de la maîtresse de ces appartements respirait tout entier dans le salon où attendait Augustine. Elle tâcha d’y deviner le caractère de sa rivale par l’aspect des objets épars ; mais il y avait là quelque chose d’impénétrable dans le désordre comme dans la symétrie, et pour la simple Augustine ce fut lettres closes. Tout ce qu’elle put y voir, c’est que la duchesse était une femme supérieure en tant que femme. Elle eut alors une pensée douloureuse.\\n\\n— Hélas ! serait-il vrai, se dit-elle, qu’un coeur aimant et simple ne suffit pas à un artiste ; et pour balancer le poids de ces âmes fortes, faut-il les unir à des âmes féminines dont la puissance soit pareille à la leur ? Si j’avais été élevée comme cette sirène, au moins nos armes eussent été égales au moment de la lutte.\\n\\n— Mais je n’y suis pas ! Ces mots secs et brefs, quoique prononcés à voix basse dans le boudoir voisin, furent entendus par Augustine, dont le coeur palpita.\\n\\n— Cette dame est là , répliqua la femme de chambre.\\n\\n— Vous êtes folle, faites donc entrer ! répondit la duchesse dont la voix devenue douce avait pris l’accent affectueux de la politesse. Evidemment, elle désirait alors être entendue.\\n\\nAugustine s’avança timidement. Au fond de ce frais boudoir elle vit la duchesse voluptueusement couchée sur une ottomane en velours vert placée au centre d’une espèce de demi-cercle dessiné par les plis moelleux d’une mousseline tendue sur un fond jaune. Des ornements de bronze doré, disposés avec un goût exquis, rehaussaient encore cette espèce de dais sous lequel la duchesse était posée comme une statue antique. La couleur foncée du velours ne lui laissait perdre aucun moyen de séduction. Un demi-jour, ami de sa beauté, semblait être plutôt un reflet qu’une lumière. Quelques fleurs rares élevaient leurs têtes embaumées au dessus des vases de Sèvres les plus riches. Au moment où ce tableau s’offrit aux yeux d’Augustine étonnée, elle avait marché si doucement, qu’elle put surprendre un regard de l’enchanteresse. Ce regard semblait dire à une personne que la femme du peintre n’aperçut pas d’abord :\\n\\n— Restez, vous allez voir une jolie femme, et vous me rendrez sa visite moins ennuyeuse.\\n\\nA l’aspect d’Augustine, la duchesse se leva et la fit asseoir auprès d’elle.\\n\\n— A quoi dois-je le bonheur de cette visite, madame ? dit-elle avec un sourire plein de grâces.\\n\\n— Pourquoi tant de fausseté ? pensa Augustine, qui ne répondit que par une inclination de tête.\\n\\nCe silence était commandé. La jeune femme voyait devant elle un témoin de trop à cette scène. Ce personnage était, de tous les colonels de l’armée, le plus jeune, le plus élégant et le mieux fait. Son costume demi-bourgeois faisait ressortir les grâces de sa personne. Sa figure pleine de vie, de jeunesse, et déjà fort expressive, était encore animée par de petites moustaches relevées en pointe et noires comme du jais, par une impériale bien fournie, par des favoris soigneusement peignés et par une forêt de cheveux noirs assez en désordre. Il badinait avec une cravache, en manifestant une aisance et une liberté qui seyaient à l’air satisfait de sa physionomie ainsi qu’à la recherche de sa toilette. Les rubans attachés à sa boutonnière étaient noués avec dédain, et il paraissait bien plus vain de sa jolie tournure que de son courage. Augustine regarda la duchesse de Carigliano en lui montrant le colonel par un coup d’oeil dont toutes les prières furent comprises.\\n\\n— Eh bien, adieu, monsieur d’Aiglemont, nous nous retrouverons au bois de Boulogne.\\n\\nCes mots furent prononcés par la sirène comme s’ils étaient le résultat d’une stipulation antérieure à l’arrivée d’Augustine ; elle les accompagna d’un regard menaçant que l’officier méritait peut-être pour l’admiration qu’il témoignait en contemplant la modeste fleur qui contrastait si bien avec l’orgueilleuse duchesse. Le jeune fat s’inclina en silence, tourna sur les talons de ses bottes, et s’élança gracieusement hors du boudoir. En ce moment, Augustine, épiant sa rivale qui semblait suivre des yeux le brillant officier, surprit dans ce regard un sentiment dont les fugitives expressions sont connues de toutes les femmes. Elle songea avec la douleur la plus profonde que sa visite allait être inutile : cette artificieuse duchesse était trop avide d’hommages pour ne pas avoir le coeur sans pitié.\\n\\n— Madame, dit Augustine d’une voix entrecoupée, la démarche que je fais en ce moment auprès de vous va vous sembler bien singulière ; mais le désespoir a sa folie, et doit faire tout excuser. Je m’explique trop bien pourquoi Théodore préfère votre maison à toute autre, et pourquoi votre esprit exerce tant d’empire sur lui. Hélas ! je n’ai qu’à rentrer en moi-même pour en trouver des raisons plus que suffisantes. Mais j’adore mon mari, madame. Deux ans de larmes n’ont point effacé son image de mon coeur, quoique j’aie perdu le sien. Dans ma folie, j’ai osé concevoir l’idée de lutter avec vous ; et je viens à vous, vous demander par quels moyens je puis triompher de vous-même. Oh, madame ! s’écria la jeune femme en saisissant avec ardeur la main de sa rivale, qui la lui laissa prendre, je ne prierai jamais Dieu pour mon propre bonheur avec autant de ferveur que je l’implorerais pour le vôtre, si vous m’aidiez à reconquérir, je ne dirai pas l’amour, mais la tendresse de Sommervieux. Je n’ai plus d’espoir qu’en vous. Ah ! dites-moi comment vous avez pu lui plaire et lui faire oublier les premiers jours de…\\n\\nA ces mots, Augustine, suffoquée par des sanglots mal contenus, fut obligée de s’arrêter. Honteuse de sa faiblesse, elle cacha son visage dans un mouchoir qu’elle inonda de ses larmes.\\n\\n— Etes-vous donc enfant, ma chère petite belle ! dit la duchesse, qui, séduite par la nouveauté de cette scène et attendrie malgré elle en recevant l’hommage que lui rendait la plus parfaite vertu qui fût peut-être à Paris, prit le mouchoir de la jeune femme et se mit à lui essuyer elle-même les yeux en la flattant par quelques monosyllabes murmurés avec une gracieuse pitié.\\n\\nAprès un moment de silence, la coquette, emprisonnant les jolies mains de la pauvre Augustine entre les siennes qui avaient un rare caractère de beauté noble et de puissance, lui dit d’une voix douce et affectueuse :\\n\\n— Pour premier avis, je vous conseillerai de ne pas pleurer ainsi, les larmes enlaidissent. Il faut savoir prendre son parti sur les chagrins ; ils rendent malade, et l’amour ne reste pas long-temps sur un lit de douleur. La mélancolie donne bien d’abord une certaine grâce qui plaît ; mais elle finit par allonger les traits et flétrir la plus ravissante de toutes les figures. Ensuite, nos tyrans ont l’amour-propre de vouloir que leurs esclaves soient toujours gaies.\\n\\n— Ah, madame ! il ne dépend pas de moi de ne pas sentir ! Comment peut-on, sans éprouver mille morts, voir terne, décolorée, indifférente, une figure qui jadis rayonnait d’amour et de joie ? Ah ! je ne sais pas commander à mon coeur.\\n\\n— Tant pis, chère belle ; mais je crois déjà savoir toute votre histoire. D’abord, imaginez-vous bien que si votre mari vous a été infidèle, je ne suis pas sa complice. Si j’ai tenu à l’avoir dans mon salon, c’est, je l’avouerai, par amour-propre : il était célèbre et n’allait nulle part. Je vous aime déjà trop pour vous dire toutes les folies qu’il a faites pour moi. Je ne vous en révélerai qu’une seule, parce qu’elle nous servira peut-être à vous le ramener et à le punir de l’audace qu’il met dans ses procédés avec moi. Il finirait par me compromettre. Je connais trop le monde, ma chère, pour vouloir me mettre à la discrétion d’un homme trop supérieur. Sachez qu’il faut se laisser faire la cour par eux, mais les épouser ! c’est une faute. Nous autres femmes, nous devons admirer les hommes de génie, en jouir comme d’un spectacle, mais vivre avec eux ! jamais. Fi donc ! c’est vouloir prendre plaisir à regarder les machines de l’opéra, au lieu de rester dans une loge, à y savourer ses brillantes illusions. Mais chez vous, ma pauvre enfant, le mal est arrivé, n’est-ce pas ? Eh bien ! il faut essayer de vous armer contre la tyrannie.\\n\\n— Ah, madame ! avant d’entrer ici, en vous y voyant, j’ai déjà reconnu quelques artifices que je ne soupçonnais pas.\\n\\n— Eh bien, venez me voir quelquefois, et vous ne serez pas long-temps sans posséder la science de ces bagatelles, d’ailleurs assez importantes. Les choses extérieures sont, pour les sots, la moitié de la vie ; et pour cela, plus d’un homme de talent se trouve un sot malgré tout son esprit. Mais je gage que vous n’avez jamais rien su refuser à Théodore ?\\n\\n— Le moyen, madame, de refuser quelque chose à celui qu’on aime !\\n\\n— Pauvre innocente, je vous adorerais pour votre niaiserie. Sachez donc que plus nous aimons, moins nous devons laisser apercevoir à un homme, surtout à un mari, l’étendue de notre passion. C’est celui qui aime le plus qui est tyrannisé, et, qui pis est, délaissé tôt ou tard. Celui qui veut régner, doit…\\n\\n— Comment, madame ! faudra-t-il donc dissimuler, calculer, devenir fausse, se faire un caractère artificiel et pour toujours ? Oh ! comment peut-on vivre ainsi ? Est-ce que vous pouvez…\\n\\nElle hésita, la duchesse sourit.\\n\\n— Ma chère, reprit la grande dame d’une voix grave, le bonheur conjugal a été de tout temps une spéculation, une affaire qui demande une attention particulière. Si vous continuez à parler passion quand je vous parle mariage, nous ne nous entendrons bientôt plus. Ecoutez-moi, continua-t-elle en prenant le ton d’une confidence. J’ai été à même de voir quelques-uns des hommes supérieurs de notre époque. Ceux qui se sont mariés ont, à quelques exceptions près, épousé des femmes nulles. Eh bien ! ces femmes-là les gouvernaient, comme l’empereur nous gouverne, et étaient, sinon aimées, du moins respectées par eux. J’aime assez les secrets, surtout ceux qui nous concernent, pour m’être amusée à chercher le mot de cette énigme. Eh bien, mon ange ! ces bonnes femmes avaient le talent d’analyser le caractère de leurs maris. Sans s’épouvanter comme vous de leurs supériorités, elles avaient adroitement remarqué les qualités qui leur manquaient. Soit qu’elles possédassent ces qualités, ou qu’elles feignissent de les avoir, elles trouvaient moyen d’en faire un si grand étalage aux yeux de leurs maris qu’elles finissaient par leur imposer. Enfin, apprenez encore que ces âmes qui paraissent si grandes ont toutes un petit grain de folie que nous devons savoir exploiter. En prenant la ferme volonté de les dominer, en ne s’écartant jamais de ce but, en y rapportant toutes nos actions, nos idées, nos coquetteries, nous maîtrisons ces esprits éminemment capricieux qui, par la mobilité même de leurs pensées, nous donnent les moyens de les influencer.\\n\\n— Oh ciel ! s’écria la jeune femme épouvantée, voilà donc la vie.\\n\\nC’est un combat…\\n\\n\\n\\n— Où il faut toujours menacer, reprit la duchesse en riant. Notre pouvoir est tout factice. Aussi ne faut-il jamais se laisser mépriser par un homme ; on ne se relève d’une pareille chute que par des manoeuvres odieuses. Venez, ajouta-t-elle, je vais vous donner un moyen de mettre votre mari à la chaîne.\\n\\nElle se leva, pour guider en souriant la jeune et innocente apprentie des ruses conjugales à travers le dédale de son petit palais. Elles arrivèrent toutes deux à un escalier dérobé qui communiquait aux appartements de réception. Quand la duchesse tourna le secret de la porte, elle s’arrêta, regarda Augustine avec un air inimitable de finesse et de grâce :\\n\\n— Tenez, le duc de Carigliano m’adore ! eh bien, il n’ose pas entrer par cette porte sans ma permission. Et c’est un homme qui a l’habitude de commander à des milliers de soldats. Il sait affronter les batteries, mais devant moi ! il a peur.\\n\\nAugustine soupira. Elles parvinrent à une somptueuse galerie où la femme du peintre fut amenée par la duchesse devant le portrait que Théodore avait fait de mademoiselle Guillaume. A cet aspect, Augustine jeta un cri.\\n\\n— Je savais bien qu’il n’était plus chez moi, dit-elle, mais… ici !\\n\\n— Ma chère, je ne l’ai exigé que pour voir jusqu’à quel degré de bêtise un homme de génie peut atteindre. Tôt ou tard, il vous aurait été rendu par moi ; mais je ne m’attendais pas au plaisir de voir ici l’original devant la copie. Pendant que nous allons achever notre conversation, je le ferai porter dans votre voiture. Si, armée de ce talisman, vous n’êtes pas maîtresse de votre mari pendant cent ans, vous n’êtes pas une femme, et vous méritez votre sort !\\n\\nAugustine baisa la main de la duchesse, qui la pressa sur son coeur et l’embrassa avec une tendresse d’autant plus vive qu’elle devait être oubliée le lendemain. Cette scène aurait peut-être à jamais ruiné la candeur et la pureté d’une femme moins vertueuse qu’Augustine, à qui les secrets révélés par la duchesse pouvaient être également salutaires et funestes. La politique astucieuse des hautes sphères sociales ne convenait pas plus à Augustine que l’étroite raison de Joseph Lebas, ou que la niaise morale de madame Guillaume. Etrange effet des fausses positions où nous jettent les moindres contresens commis dans la vie ! Augustine ressemblait alors à un pâtre des Alpes surpris par une avalanche : s’il hésite, ou s’il veut écouter les cris de ses compagnons, le plus souvent il périt. Dans ces grandes crises, le coeur se brise ou se bronze.\\n\\nMadame de Sommervieux revint chez elle en proie à une agitation qu’il serait difficile de décrire. Sa conversation avec la duchesse de Carigliano éveillait une foule d’idées contradictoires dans son esprit. Elle était comme les moutons de la fable, pleine de courage en l’absence du loup. Elle se haranguait elle-même et se traçait d’admirables plans de conduite ; elle concevait mille stratagèmes de coquetterie ; elle parlait même à son mari, retrouvant, loin de lui, toutes les ressources de cette éloquence vraie qui n’abandonne jamais les femmes ; puis, en songeant au regard fixe et clair de Théodore, elle tremblait déjà . Quand elle demanda si monsieur était chez lui, la voix lui manqua. En apprenant qu’il ne reviendrait pas dîner, elle éprouva un mouvement de joie inexplicable. Semblable au criminel qui se pourvoit en cassation contre son arrêt de mort, un délai, quelque court qu’il pût être, lui semblait une vie entière. Elle plaça le portrait dans sa chambre, et attendit son mari en se livrant à toutes les angoisses de l’espérance Elle pressentait trop bien que cette tentative allait décider de tout son avenir, pour ne pas frissonner à toute espèce de bruit, même au murmure de sa pendule qui semblait appesantir ses terreurs en les lui mesurant. Elle tâcha de tromper le temps par mille artifices. Elle eut l’idée de faire une toilette qui la rendit semblable en tout point au portrait. Puis, connaissant le caractère inquiet de son mari, elle fit éclairer son appartement d’une manière inusitée, certaine qu’en rentrant la curiosité l’amènerait chez elle. Minuit sonna, quand, au cri du jockey, la porte de l’hôtel s’ouvrit. La voiture du peintre roula sur le pavé de la cour silencieuse.\\n\\n— Que signifie cette illumination ? demanda Théodore d’une voix joyeuse en entrant dans la chambre de sa femme.\\n\\nAugustine saisit avec adresse un moment si favorable, elle s’élança au cou de son mari et lui montra le portrait. L’artiste resta immobile comme un rocher. Ses yeux se dirigèrent alternativement sur Augustine et sur la toile accusatrice. La timide épouse, demi-morte, épiait le front changeant, le front terrible de son mari. Elle en vit par degrés les rides expressives s’amonceler comme des nuages ; puis, elle crut sentir son sang se figer dans ses veines, quand, par un regard flamboyant et d’une voix profondément sourde, elle fut interrogée.\\n\\n— Où avez-vous trouvé ce tableau ?\\n\\n— La duchesse de Carigliano me l’a rendu.\\n\\n— Vous le lui avez demandé ?\\n\\n— Je ne savais pas qu’il fût chez elle.\\n\\nLa douceur ou plutôt la mélodie enchanteresse de la voix de cet ange eût attendri des Cannibales, mais non un artiste en proie aux tortures de la vanité blessée.\\n\\n— Cela est digne d’elle, s’écria l’artiste d’une voix tonnante. Je me vengerai ! dit-il en se promenant à grands pas. Elle en mourra de honte : je la peindrai ! oui, je la représenterai sous les traits de Messaline sortant à la nuit du palais de Claude.\\n\\n— Théodore ! dit une voix mourante.\\n\\n— Je la tuerai.\\n\\n— Mon ami !\\n\\n— Elle aime ce petit colonel de cavalerie, parce qu’il monte bien à cheval…\\n\\n— Théodore !\\n\\n— Eh ! laissez-moi, dit le peintre à sa femme avec un son de voix qui ressemblait presque à un rugissement.\\n\\nIl serait odieux de peindre toute cette scène à la fin de laquelle l’ivresse de la colère suggéra à l’artiste des paroles et des actes qu’une femme, moins jeune qu’Augustine, aurait attribués à la démence.\\n\\nSur les huit heures du matin, le lendemain, madame Guillaume surprit sa fille pâle, les yeux rouges, la coiffure en désordre, tenant à la main un mouchoir trempé de pleurs, contemplant sur le parquet les fragments épars d’une toile déchirée et les morceaux d’un grand cadre doré mis en pièce. Augustine, que la douleur rendait presque insensible, montra ces débris par un geste empreint de désespoir.\\n\\n— Et voilà peut-être une grande perte, s’écria la vieille régente du Chat-qui-pelote. Il était ressemblant, c’est vrai ; mais j’ai appris qu’il y a sur le boulevard un homme qui fait des portraits charmants pour cinquante écus.\\n\\n— Ah, ma mère !\\n\\n— Pauvre petite, tu as bien raison ! répondit madame Guillaume qui méconnut l’expression du regard que lui jeta sa fille. Va, mon enfant, l’on n’est jamais si tendrement aimé que par sa mère. Ma mignonne, je devine tout ; mais viens me confier tes chagrins, je te consolerai. Ne t’ai-je pas déjà dit que cet homme-là était un fou ! Ta femme de chambre m’a conté de belles choses… Mais c’est donc un véritable monstre !\\n\\nAugustine mit un doigt sur ses lèvres pâlies, comme pour implorer de sa mère un moment de silence. Pendant cette terrible nuit, le malheur lui avait fait trouver cette patiente résignation qui, chez les mères et chez les femmes aimantes, surpasse, dans ses effets, l’énergie humaine et révèle peut-être dans le coeur des femmes l’existence de certaines cordes que Dieu a refusées à l’homme.\\n\\nUne inscription gravée sur un cippe du cimetière Montmartre indiquait que madame de Sommervieux était morte à vingt-sept ans. Un poète, ami de cette timide créature, voyait, dans les simples lignes de son épitaphe, la dernière scène d’un drame. Chaque année, au jour solennel du 2 novembre, il ne passait jamais devant ce jeune marbre sans se demander s’il ne fallait pas des femmes plus fortes que ne l’était Augustine pour les puissantes étreintes du génie.\\n\\n— Les humbles et modestes fleurs, écloses dans les vallées, meurent peut-être, se disait-il, quand elles sont transplantées trop près des cieux, aux régions où se forment les orages, où le soleil est brûlant.\\n\\nMaffliers, octobre 1829.\\n\\n\\n\\n'"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 54
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "6zlbrIDYw_cO"
+ },
+ "source": [
+ "## Transform novels to sentence lists"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "A_cSaehyj23t",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 72
+ },
+ "outputId": "364f2b80-5631-4b42-b4e0-4dbe21176ca0"
+ },
+ "source": [
+ "import nltk\n",
+ "nltk.download('punkt')\n"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
+ "[nltk_data] Unzipping tokenizers/punkt.zip.\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "True"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 55
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "zHtGN7EAvBkG"
+ },
+ "source": [
+ "#Function for split the sentences of text using \"re\" library\n",
+ "import re\n",
+ "alphabets= \"([A-Za-z])\"\n",
+ "prefixes = \"(Mr|St|Mme|Mlle|Mrs|Ms|Dr)[.]\"\n",
+ "suffixes = \"(Inc|Ltd|Jr|Sr|Co)\"\n",
+ "starters = \"(Mr|St|Mme|Mlle|Mrs|Ms|Dr|Il\\s|Elle\\s|It\\s|Ils\\s|Elle\\s|Leur\\s|Notre\\s|Nous\\s|On\\s|Mais\\s|Cependant\\s|Ce\\s|Cette\\s|He\\s|She\\s|It\\s|They\\s|Their\\s|Our\\s|We\\s|But\\s|However\\s|That\\s|This\\s|Wherever)\"\n",
+ "acronyms = \"([A-Z][.][A-Z][.](?:[A-Z][.])?)\"\n",
+ "websites = \"[.](com|net|org|io|gov)\"\n",
+ "\n",
+ "def split_into_sentences(text):\n",
+ " text = \" \" + text + \" \"\n",
+ " text = text.replace(\"\\n\",\" \")\n",
+ " text = re.sub(prefixes,\"\\\\1<prd>\",text)\n",
+ " text = re.sub(websites,\"<prd>\\\\1\",text)\n",
+ " if \"Ph.D\" in text: text = text.replace(\"Ph.D.\",\"Ph<prd>D<prd>\")\n",
+ " text = re.sub(\"\\s\" + alphabets + \"[.] \",\" \\\\1<prd> \",text)\n",
+ " text = re.sub(acronyms+\" \"+starters,\"\\\\1<stop> \\\\2\",text)\n",
+ " text = re.sub(alphabets + \"[.]\" + alphabets + \"[.]\" + alphabets + \"[.]\",\"\\\\1<prd>\\\\2<prd>\\\\3<prd>\",text)\n",
+ " text = re.sub(alphabets + \"[.]\" + alphabets + \"[.]\",\"\\\\1<prd>\\\\2<prd>\",text)\n",
+ " text = re.sub(\" \"+suffixes+\"[.] \"+starters,\" \\\\1<stop> \\\\2\",text)\n",
+ " text = re.sub(\" \"+suffixes+\"[.]\",\" \\\\1<prd>\",text)\n",
+ " text = re.sub(\" \" + alphabets + \"[.]\",\" \\\\1<prd>\",text)\n",
+ " #..... -> .\n",
+ " # text = re.sub('[.]+', '.', text)\n",
+ " if \"â€\" in text: text = text.replace(\".â€\",\"â€.\")\n",
+ " if \"\\\"\" in text: text = text.replace(\".\\\"\",\"\\\".\")\n",
+ " if \"!\" in text: text = text.replace(\"!\\\"\",\"\\\"!\")\n",
+ " if \"?\" in text: text = text.replace(\"?\\\"\",\"\\\"?\")\n",
+ " text = text.replace(\".\",\".<stop>\")\n",
+ " text = text.replace(\"?\",\"?<stop>\")\n",
+ " text = text.replace(\"!\",\"!<stop>\")\n",
+ " text = text.replace(\"<prd>\",\".\")\n",
+ " sentences = text.split(\"<stop>\")\n",
+ " sentences = sentences[:-1]\n",
+ " sentences = [s.strip() for s in sentences]\n",
+ " return sentences"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "t6B3rVVxGPPC"
+ },
+ "source": [
+ "splited_sentences= [None] * len(content_french)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "aUrnnnRJ6I6J"
+ },
+ "source": [
+ "for i in range(len(content_french)):\n",
+ " splited_sentences[i]=split_into_sentences(content_french[i])\n",
+ "\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "HkZ1GMANQfY5",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "584f765c-b895-4401-de02-86d5dc1b8252"
+ },
+ "source": [
+ "for i in range(len(splited_sentences)):\n",
+ " print(\"nombre de phrases du roman \" + onlyfiles[i], len(splited_sentences[i]))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "nombre de phrases du roman maison.txt 1010\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "9p_pe4i4N2nN"
+ },
+ "source": [
+ "test_sentences = [None] * len(splited_sentences)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Wec_cqhf0a3v"
+ },
+ "source": [
+ "#Convert the lists of sentences to ndarray by numpy\n",
+ "for i in range(len(splited_sentences)):\n",
+ " test_sentences[i] = np.asarray(splited_sentences[i], dtype=np.str)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "tzGxUSXE2act",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "c3d50004-8e82-4226-ee50-93951d030b02"
+ },
+ "source": [
+ "onlyfiles[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "'maison.txt'"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 63
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "baooh9Bs1a6U"
+ },
+ "source": [
+ "import pandas as pd\n",
+ "\n",
+ "#input_ids_test = [None] * len(test_sentences)\n",
+ "#attention_masks = [None] * len()\n",
+ "prediction_dataloader = [None] * len(test_sentences)\n",
+ "\n",
+ "for i in range(len(test_sentences)):\n",
+ " # Tokenize all of the sentences and map the tokens to thier word IDs.\n",
+ " #input_ids_test[i] = [None] * len(test_sentences[i])\n",
+ " #prediction_dataloader[i] = [None] * len(test_sentences[i])\n",
+ " input_ids_test = []\n",
+ " # For every sentence...\n",
+ " for sent in test_sentences[i]:\n",
+ " # `encode` will:\n",
+ " # (1) Tokenize the sentence.\n",
+ " # (2) Prepend the `[CLS]` token to the start.\n",
+ " # (3) Append the `[SEP]` token to the end.\n",
+ " # (4) Map tokens to their IDs.\n",
+ " encoded_sent = tokenizer.encode(\n",
+ " sent, # Sentence to encode.\n",
+ " add_special_tokens = True, # Add '[CLS]' and '[SEP]'\n",
+ " )\n",
+ " \n",
+ " input_ids_test.append(encoded_sent)\n",
+ "\n",
+ " # Pad our input tokens\n",
+ " input_ids_test = pad_sequences(input_ids_test, maxlen=MAX_LEN, \n",
+ " dtype=\"long\", truncating=\"post\", padding=\"post\")\n",
+ "\n",
+ " # Create attention masks\n",
+ " attention_masks = []\n",
+ "\n",
+ " # Create a mask of 1s for each token followed by 0s for padding\n",
+ " for seq in input_ids_test:\n",
+ " seq_mask = [float(i>0) for i in seq]\n",
+ " attention_masks.append(seq_mask) \n",
+ "\n",
+ " # Convert to tensors.\n",
+ " prediction_inputs = torch.tensor(input_ids_test)\n",
+ " prediction_masks = torch.tensor(attention_masks)\n",
+ " #prediction_labels = torch.tensor(labels)\n",
+ "\n",
+ " # Set the batch size. \n",
+ " batch_size = 32 \n",
+ "\n",
+ " # Create the DataLoader.\n",
+ " prediction_data = TensorDataset(prediction_inputs, prediction_masks)\n",
+ " prediction_sampler = SequentialSampler(prediction_data)\n",
+ " prediction_dataloader[i] = DataLoader(prediction_data, sampler=prediction_sampler, batch_size=batch_size)\n",
+ "\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Xc8fZs081IhG",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "c5dd8217-68ff-48e1-da15-ae613da993d3"
+ },
+ "source": [
+ "predictions = []\n",
+ "# Prediction on novels\n",
+ "for i in range(len(prediction_dataloader)):\n",
+ " print('Predicting labels for {:,} test sentences '.format(len(test_sentences[i])) + 'for the novel ' + onlyfiles[i])\n",
+ "\n",
+ " # Put model in evaluation mode\n",
+ " model.eval()\n",
+ "\n",
+ " # Tracking variables \n",
+ " predictions_i = []\n",
+ "\n",
+ " # Predict \n",
+ " for batch in prediction_dataloader[i]:\n",
+ " # Add batch to GPU\n",
+ " batch = tuple(t.to(device) for t in batch)\n",
+ " \n",
+ " # Unpack the inputs from our dataloader\n",
+ " b_input_ids, b_input_mask= batch\n",
+ " \n",
+ " # Telling the model not to compute or store gradients, saving memory and \n",
+ " # speeding up prediction\n",
+ " with torch.no_grad():\n",
+ " # Forward pass, calculate logit predictions\n",
+ " outputs = model(b_input_ids, token_type_ids=None, \n",
+ " attention_mask=b_input_mask)\n",
+ " #print(outputs[0])\n",
+ " logits = outputs[0]\n",
+ "\n",
+ " # Move logits and labels to CPU\n",
+ " logits = logits.detach().cpu().numpy()\n",
+ " #label_ids = b_labels.to('cpu').numpy()\n",
+ " \n",
+ " # Store predictions and true labels\n",
+ " predictions_i.append(logits)\n",
+ " #print(predictions_i)\n",
+ " predictions.append(predictions_i)\n",
+ " #true_labels.append(label_ids)\n",
+ "\n",
+ "print(' DONE.')"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Predicting labels for 1,010 test sentences for the novel maison.txt\n",
+ " DONE.\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "qpA2QBQ02TYg",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "d855d2e8-a565-4ef4-f8b5-d5553a94de47"
+ },
+ "source": [
+ "pred_labels = []\n",
+ "\n",
+ "for j in range(len(prediction_dataloader)): \n",
+ " # Evaluate each test batch using Matthew's correlation coefficient\n",
+ " print('Calculating Matthews Corr. Coef. for each batch of ' + onlyfiles[j])\n",
+ " pred_labels_i= []\n",
+ " # For each input batch...\n",
+ " for i in range(len(prediction_dataloader[j])):\n",
+ " # The predictions for this batch are a 2-column ndarray (one column for \"0\" \n",
+ " # and one column for \"1\"). Pick the label with the highest value and turn this\n",
+ " # in to a list of 0s and 1s.\n",
+ " pred_labels_i_j = np.argmax(predictions[j][i], axis=1).flatten()\n",
+ " pred_labels_i.append(pred_labels_i_j)\n",
+ " pred_labels.append(pred_labels_i)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Calculating Matthews Corr. Coef. for each batch of maison.txt\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "yjNZgkLe9EOr"
+ },
+ "source": [
+ "pred_rom_list = []\n",
+ "for pred_rom in range(len(pred_labels)):\n",
+ " pred_rom_i_list = []\n",
+ " for pred_bat in range(len(pred_labels[pred_rom])):\n",
+ " pred_rom_i_list.append(pred_labels[pred_rom][pred_bat].tolist())\n",
+ " pred_rom_list.append(pred_rom_i_list)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "jOb1CQRTCWKW"
+ },
+ "source": [
+ "roms_list_fin= []\n",
+ "for i in range(len(pred_rom_list)):\n",
+ " flat_list = [item for sublist in pred_rom_list[i] for item in sublist]\n",
+ " roms_list_fin.append(flat_list)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "u43hq4sjGokV"
+ },
+ "source": [
+ "# on va voir dans les résultats parfois 1 pour une phrase qui n'indique pas sentences Geo.\n",
+ "# mais cette phrase est liée soit à une phrase Geo avant soit à une phrase Geo après\n",
+ "import pandas\n",
+ "for i in range(len(onlyfiles)):\n",
+ " df = pandas.DataFrame(data={\"sentences\": test_sentences[i], \"labels\": roms_list_fin[i]})\n",
+ " df.to_csv(\"./resultats_\" + onlyfiles[i] + \".csv\", sep=',',index=False)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "jAU0kLCAjPme"
+ },
+ "source": [
+ "# Using Bert uncased"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Sq-8aExPjKMS"
+ },
+ "source": [
+ "# 1 Loading Dataset\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "lNPC8JB3jKMT"
+ },
+ "source": [
+ "## 1.1. Download & Extract"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "fOUIHY3WjKMT"
+ },
+ "source": [
+ "import pandas as pd"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "TALdq9y9jKMV",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "306b7661-49ac-4cfa-d97c-66d485230d5f"
+ },
+ "source": [
+ "Geo_analyse = pd.read_csv(\"/content/train_moins_sample.csv\")\n",
+ "\n",
+ "Geo_analyse.isnull().values.any()\n",
+ "\n",
+ "Geo_analyse.shape"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "(2317, 2)"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 73
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "WJ_l4NLyjKMY"
+ },
+ "source": [
+ "## 1.2. Parse"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "1qDz8BJ2jKMY"
+ },
+ "source": [
+ "Here are five sentences of the dataset"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "KakSZF6ujKMa",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 206
+ },
+ "outputId": "854055b6-3019-4200-be97-c6f013c042ef"
+ },
+ "source": [
+ "Geo_analyse.head(5)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>sentences</th>\n",
+ " <th>labels</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>— Comme tu voudras, répondit-elle.</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>Ce n’est plus moi, c’est elle qui couche avec ...</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>Un parent, je crois.</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>je m’y attendais un peu.</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>Comment la faire taire?</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " sentences labels\n",
+ "0 — Comme tu voudras, répondit-elle. 0\n",
+ "1 Ce n’est plus moi, c’est elle qui couche avec ... 0\n",
+ "2 Un parent, je crois. 0\n",
+ "3 je m’y attendais un peu. 0\n",
+ "4 Comment la faire taire? 0"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 74
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "eDKYsZLnjKMg"
+ },
+ "source": [
+ "\n",
+ "\n",
+ "Let's extract the sentences and labels of our training set as numpy ndarrays."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "ng6fXXhDjKMh"
+ },
+ "source": [
+ "# Get the lists of sentences and their labels.\n",
+ "sentences = Geo_analyse.sentences.values\n",
+ "labels = Geo_analyse.labels.values"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "OUvzEhXXC8vE"
+ },
+ "source": [
+ "#2 Tokenization and Input Formating"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "nbbSfa32jKMj"
+ },
+ "source": [
+ "## 2.1. BERT Tokenizer (uncased)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "qrpfTBeWjKMl",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 86,
+ "referenced_widgets": [
+ "ae4a115927af4738b9c3f2a8cb97a32d",
+ "da83613435044dfdb693f7cd6293898e",
+ "6a0aa93d9b3841d7934cd07768b4e8d9",
+ "3fd2e1b55945449e81a96e89b5e0015d",
+ "7f7f5353878d4046b5a1e0b831f923b9",
+ "49650cd4a0664497839d492c0911acc7",
+ "28aef9ee16b24d4d94ff728d7e7f7b6a",
+ "b8f06024ff9f4d24aa087732be6baeab"
+ ]
+ },
+ "outputId": "6cf22b11-7563-47a5-d312-4ae7603fc821"
+ },
+ "source": [
+ "from transformers import BertTokenizer\n",
+ "\n",
+ "# Load the BERT tokenizer.\n",
+ "print('Loading BERT tokenizer...')\n",
+ "tokenizer = BertTokenizer.from_pretrained('bert-base-multilingual-uncased',do_lower_case=True)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Loading BERT tokenizer...\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "ae4a115927af4738b9c3f2a8cb97a32d",
+ "version_minor": 0,
+ "version_major": 2
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=871891.0, style=ProgressStyle(descripti…"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Bl8MrPYbjKMm"
+ },
+ "source": [
+ "Let's apply the tokenizer to one sentence just to see the output.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "f9KG73M1jKMn",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 73
+ },
+ "outputId": "f27af934-1e50-4d8a-8eab-662ece6e48cc"
+ },
+ "source": [
+ "# Print the original sentence.\n",
+ "print(' Original: ', sentences[0])\n",
+ "\n",
+ "# Print the sentence split into tokens.\n",
+ "print('Tokenized: ', tokenizer.tokenize(sentences[0]))\n",
+ "\n",
+ "# Print the sentence mapped to token ids.\n",
+ "print('Token IDs: ', tokenizer.convert_tokens_to_ids(tokenizer.tokenize(sentences[0])))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ " Original: — Comme tu voudras, répondit-elle.\n",
+ "Tokenized: ['[UNK]', 'comme', 'tu', 'vo', '##udra', '##s', ',', 'rep', '##ond', '##it', '-', 'elle', '.']\n",
+ "Token IDs: [100, 11043, 10689, 11821, 94698, 10107, 117, 37090, 21510, 10517, 118, 10725, 119]\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "UwYOcd33jKMt"
+ },
+ "source": [
+ "## 2.2. Sentences to IDs"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "q7qIIgzajKMu",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "0e568c97-ace3-4123-e29e-7d272092bddd"
+ },
+ "source": [
+ "# Tokenize all of the sentences and map the tokens to thier word IDs.\n",
+ "input_ids = []\n",
+ "\n",
+ "# For every sentence...\n",
+ "for sent in sentences:\n",
+ " # `encode` will:\n",
+ " # (1) Tokenize the sentence.\n",
+ " # (2) Prepend the `[CLS]` token to the start.\n",
+ " # (3) Append the `[SEP]` token to the end.\n",
+ " # (4) Map tokens to their IDs.\n",
+ " encoded_sent = tokenizer.encode(\n",
+ " sent, # Sentence to encode.\n",
+ " add_special_tokens = True, # Add '[CLS]' and '[SEP]'\n",
+ "\n",
+ " # This function also supports truncation and conversion\n",
+ " # to pytorch tensors, but we need to do padding, so we\n",
+ " # can't use these features :( .\n",
+ " #max_length = 128, # Truncate all sentences.\n",
+ " #return_tensors = 'pt', # Return pytorch tensors.\n",
+ " )\n",
+ " \n",
+ " # Add the encoded sentence to the list.\n",
+ " input_ids.append(encoded_sent)\n",
+ "\n",
+ "# Print sentence 0, now as a list of IDs.\n",
+ "print('Original: ', sentences[0])\n",
+ "print('Token IDs:', input_ids[0])"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Original: — Comme tu voudras, répondit-elle.\n",
+ "Token IDs: [101, 100, 11043, 10689, 11821, 94698, 10107, 117, 37090, 21510, 10517, 118, 10725, 119, 102]\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "d2afWRPjjKMv"
+ },
+ "source": [
+ "## 2.3. Padding & Truncating"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "oV25Rn_6jKMw",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "5d9f35dd-e729-463e-8dae-3f39e417416d"
+ },
+ "source": [
+ "print('Max sentence length: ', max([len(sen) for sen in input_ids]))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Max sentence length: 232\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "2V1HJKIljKMz"
+ },
+ "source": [
+ "Given that, let's choose MAX_LEN = 256 and apply the padding."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "i4xZqZQpjKMz",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 129
+ },
+ "outputId": "01963a14-28f4-42c0-fe04-22de8e1e8e32"
+ },
+ "source": [
+ "# We'll borrow the `pad_sequences` utility function to do this.\n",
+ "from keras.preprocessing.sequence import pad_sequences\n",
+ "\n",
+ "# Set the maximum sequence length.\n",
+ "# I've chosen 256 somewhat arbitrarily. It's slightly larger than the\n",
+ "# maximum training sentence length of 205...\n",
+ "MAX_LEN = 256\n",
+ "\n",
+ "print('\\nPadding/truncating all sentences to %d values...' % MAX_LEN)\n",
+ "\n",
+ "print('\\nPadding token: \"{:}\", ID: {:}'.format(tokenizer.pad_token, tokenizer.pad_token_id))\n",
+ "\n",
+ "# Pad our input tokens with value 0.\n",
+ "# \"post\" indicates that we want to pad and truncate at the end of the sequence,\n",
+ "# as opposed to the beginning.\n",
+ "input_ids = pad_sequences(input_ids, maxlen=MAX_LEN, dtype=\"long\", \n",
+ " value=0, truncating=\"post\", padding=\"post\")\n",
+ "\n",
+ "print('\\nDone.')"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "\n",
+ "Padding/truncating all sentences to 256 values...\n",
+ "\n",
+ "Padding token: \"[PAD]\", ID: 0\n",
+ "\n",
+ "Done.\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "CnYrkC3ZjKM1"
+ },
+ "source": [
+ "## 2.4. Attention Masks"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "5Q4mVJNBjKM1"
+ },
+ "source": [
+ "# Create attention masks\n",
+ "attention_masks = []\n",
+ "\n",
+ "# For each sentence...\n",
+ "for sent in input_ids:\n",
+ " \n",
+ " # Create the attention mask.\n",
+ " # - If a token ID is 0, then it's padding, set the mask to 0.\n",
+ " # - If a token ID is > 0, then it's a real token, set the mask to 1.\n",
+ " att_mask = [int(token_id > 0) for token_id in sent]\n",
+ " \n",
+ " # Store the attention mask for this sentence.\n",
+ " attention_masks.append(att_mask)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "4T52cNDCjKM3"
+ },
+ "source": [
+ "## 3.5. Training & Validation Split\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "2qDYlwD2jKM4"
+ },
+ "source": [
+ "# Use train_test_split to split our data into train and validation sets for\n",
+ "# training\n",
+ "from sklearn.model_selection import train_test_split\n",
+ "\n",
+ "# Use 90% for training and 10% for validation.\n",
+ "train_inputs, validation_inputs, train_labels, validation_labels = train_test_split(input_ids, labels, \n",
+ " random_state=2018, test_size=0.1)\n",
+ "# Do the same for the masks.\n",
+ "train_masks, validation_masks, _, _ = train_test_split(attention_masks, labels,\n",
+ " random_state=2018, test_size=0.1)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "ZlL84AVOjKM5"
+ },
+ "source": [
+ "## 3.6. Converting to PyTorch Data Types"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "0mPu8eIPjKM6"
+ },
+ "source": [
+ "# Convert all inputs and labels into torch tensors, the required datatype \n",
+ "# for our model.\n",
+ "train_inputs = torch.tensor(train_inputs)\n",
+ "validation_inputs = torch.tensor(validation_inputs)\n",
+ "\n",
+ "train_labels = torch.tensor(train_labels)\n",
+ "validation_labels = torch.tensor(validation_labels)\n",
+ "\n",
+ "train_masks = torch.tensor(train_masks)\n",
+ "validation_masks = torch.tensor(validation_masks)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "E-JFZwNAjKM7"
+ },
+ "source": [
+ "from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler\n",
+ "\n",
+ "# The DataLoader needs to know our batch size for training, so we specify it \n",
+ "# here.\n",
+ "# For fine-tuning BERT on a specific task, the authors recommend a batch size of\n",
+ "# 16 or 32.\n",
+ "\n",
+ "batch_size = 32\n",
+ "\n",
+ "# Create the DataLoader for our training set.\n",
+ "train_data = TensorDataset(train_inputs, train_masks, train_labels)\n",
+ "train_sampler = RandomSampler(train_data)\n",
+ "train_dataloader = DataLoader(train_data, sampler=train_sampler, batch_size=batch_size)\n",
+ "\n",
+ "# Create the DataLoader for our validation set.\n",
+ "validation_data = TensorDataset(validation_inputs, validation_masks, validation_labels)\n",
+ "validation_sampler = SequentialSampler(validation_data)\n",
+ "validation_dataloader = DataLoader(validation_data, sampler=validation_sampler, batch_size=batch_size)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "jxNzozUejKM-"
+ },
+ "source": [
+ "# 3. Train Classification Model"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "I6hbjt1FjKNA"
+ },
+ "source": [
+ "## 3.1. BertForSequenceClassification"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "WVe_RhVVjKNB",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000,
+ "referenced_widgets": [
+ "8493b7b6fefb4462bb064057c2bbdefc",
+ "9a1ff0584fa045dc82c15ca21977581a",
+ "1030b0c14bdd4bdbb79d7cd3a04a192a",
+ "167283864c1543c7be6adb2bd0434a02",
+ "3a77f366f755438f933481739e8d4e2d",
+ "bb62a9b6ee864446a1f3c16b07e04420",
+ "b60dbc236d89455ca868903951919194",
+ "66c416fdc3774f948ff0e1ea814eaa67",
+ "a89d900918634fbd84db67e6c0eacbd3",
+ "fd849fa776164444b3d1d1c9f0a5ecff",
+ "487f4e58a4634112b021276556118d49",
+ "2156a8fec63848b285dff96898bc84d8",
+ "d906ff1cb67040c0b91d7928e1b123c4",
+ "81bf0d3185c340218382cab00eb91fe5",
+ "56626b5f15df4a9492424e20b7f31012",
+ "939183280d4a48cb8825073ece64268d"
+ ]
+ },
+ "outputId": "4d4718ed-1b63-4880-ab1a-bac26f2adedf"
+ },
+ "source": [
+ "from transformers import BertForSequenceClassification, AdamW, BertConfig\n",
+ "\n",
+ "# Load BertForSequenceClassification, the pretrained BERT model with a single \n",
+ "# linear classification layer on top. \n",
+ "model = BertForSequenceClassification.from_pretrained(\n",
+ " \"bert-base-multilingual-uncased\", # Use the 12-layer BERT model, with an uncased vocab.\n",
+ " num_labels = 2, # The number of output labels--2 for binary classification.\n",
+ " # You can increase this for multi-class tasks. \n",
+ " output_attentions = False, # Whether the model returns attentions weights.\n",
+ " output_hidden_states = False, # Whether the model returns all hidden-states.\n",
+ ")\n",
+ "\n",
+ "# Tell pytorch to run this model on the GPU.\n",
+ "model.cuda()"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "8493b7b6fefb4462bb064057c2bbdefc",
+ "version_minor": 0,
+ "version_major": 2
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=625.0, style=ProgressStyle(description_…"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "a89d900918634fbd84db67e6c0eacbd3",
+ "version_minor": 0,
+ "version_major": 2
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=672271273.0, style=ProgressStyle(descri…"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "BertForSequenceClassification(\n",
+ " (bert): BertModel(\n",
+ " (embeddings): BertEmbeddings(\n",
+ " (word_embeddings): Embedding(105879, 768, padding_idx=0)\n",
+ " (position_embeddings): Embedding(512, 768)\n",
+ " (token_type_embeddings): Embedding(2, 768)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (encoder): BertEncoder(\n",
+ " (layer): ModuleList(\n",
+ " (0): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (1): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (2): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (3): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (4): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (5): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (6): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (7): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (8): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (9): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (10): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (11): BertLayer(\n",
+ " (attention): BertAttention(\n",
+ " (self): BertSelfAttention(\n",
+ " (query): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (key): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (value): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " (output): BertSelfOutput(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " (intermediate): BertIntermediate(\n",
+ " (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
+ " )\n",
+ " (output): BertOutput(\n",
+ " (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
+ " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " )\n",
+ " )\n",
+ " )\n",
+ " )\n",
+ " (pooler): BertPooler(\n",
+ " (dense): Linear(in_features=768, out_features=768, bias=True)\n",
+ " (activation): Tanh()\n",
+ " )\n",
+ " )\n",
+ " (dropout): Dropout(p=0.1, inplace=False)\n",
+ " (classifier): Linear(in_features=768, out_features=2, bias=True)\n",
+ ")"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 85
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "sJsyNbFkjKND",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 671
+ },
+ "outputId": "201a469c-2dd4-4b3e-9290-8e5cd3bdce91"
+ },
+ "source": [
+ "# Get all of the model's parameters as a list of tuples.\n",
+ "params = list(model.named_parameters())\n",
+ "\n",
+ "print('The BERT model has {:} different named parameters.\\n'.format(len(params)))\n",
+ "\n",
+ "print('==== Embedding Layer ====\\n')\n",
+ "\n",
+ "for p in params[0:5]:\n",
+ " print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))\n",
+ "\n",
+ "print('\\n==== First Transformer ====\\n')\n",
+ "\n",
+ "for p in params[5:21]:\n",
+ " print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))\n",
+ "\n",
+ "print('\\n==== Output Layer ====\\n')\n",
+ "\n",
+ "for p in params[-4:]:\n",
+ " print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "The BERT model has 201 different named parameters.\n",
+ "\n",
+ "==== Embedding Layer ====\n",
+ "\n",
+ "bert.embeddings.word_embeddings.weight (105879, 768)\n",
+ "bert.embeddings.position_embeddings.weight (512, 768)\n",
+ "bert.embeddings.token_type_embeddings.weight (2, 768)\n",
+ "bert.embeddings.LayerNorm.weight (768,)\n",
+ "bert.embeddings.LayerNorm.bias (768,)\n",
+ "\n",
+ "==== First Transformer ====\n",
+ "\n",
+ "bert.encoder.layer.0.attention.self.query.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.self.query.bias (768,)\n",
+ "bert.encoder.layer.0.attention.self.key.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.self.key.bias (768,)\n",
+ "bert.encoder.layer.0.attention.self.value.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.self.value.bias (768,)\n",
+ "bert.encoder.layer.0.attention.output.dense.weight (768, 768)\n",
+ "bert.encoder.layer.0.attention.output.dense.bias (768,)\n",
+ "bert.encoder.layer.0.attention.output.LayerNorm.weight (768,)\n",
+ "bert.encoder.layer.0.attention.output.LayerNorm.bias (768,)\n",
+ "bert.encoder.layer.0.intermediate.dense.weight (3072, 768)\n",
+ "bert.encoder.layer.0.intermediate.dense.bias (3072,)\n",
+ "bert.encoder.layer.0.output.dense.weight (768, 3072)\n",
+ "bert.encoder.layer.0.output.dense.bias (768,)\n",
+ "bert.encoder.layer.0.output.LayerNorm.weight (768,)\n",
+ "bert.encoder.layer.0.output.LayerNorm.bias (768,)\n",
+ "\n",
+ "==== Output Layer ====\n",
+ "\n",
+ "bert.pooler.dense.weight (768, 768)\n",
+ "bert.pooler.dense.bias (768,)\n",
+ "classifier.weight (2, 768)\n",
+ "classifier.bias (2,)\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Ql7QcyDhjKNE"
+ },
+ "source": [
+ "## 3.2. Optimizer & Learning Rate Scheduler"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "OwME0RpEjKNF"
+ },
+ "source": [
+ "# Note: AdamW is a class from the huggingface library (as opposed to pytorch) \n",
+ "# I believe the 'W' stands for 'Weight Decay fix\"\n",
+ "optimizer = AdamW(model.parameters(),\n",
+ " lr = 2e-5, # args.learning_rate - default is 5e-5, our notebook had 2e-5\n",
+ " eps = 1e-8 # args.adam_epsilon - default is 1e-8.\n",
+ " )\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "fwn8y1OxjKNG"
+ },
+ "source": [
+ "from transformers import get_linear_schedule_with_warmup\n",
+ "\n",
+ "# Number of training epochs (authors recommend between 2 and 4)\n",
+ "epochs = 4\n",
+ "\n",
+ "# Total number of training steps is number of batches * number of epochs.\n",
+ "total_steps = len(train_dataloader) * epochs\n",
+ "\n",
+ "# Create the learning rate scheduler.\n",
+ "scheduler = get_linear_schedule_with_warmup(optimizer, \n",
+ " num_warmup_steps = 0, # Default value in run_glue.py\n",
+ " num_training_steps = total_steps)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "MrrkG604jKNI"
+ },
+ "source": [
+ "## 3.3. Training Loop"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "dZbne1mojKNJ"
+ },
+ "source": [
+ "Define a helper function for calculating accuracy."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "caBJqmOpjKNJ"
+ },
+ "source": [
+ "import numpy as np\n",
+ "\n",
+ "# Function to calculate the accuracy of our predictions vs labels\n",
+ "def flat_accuracy(preds, labels):\n",
+ " pred_flat = np.argmax(preds, axis=1).flatten()\n",
+ " labels_flat = labels.flatten()\n",
+ " return np.sum(pred_flat == labels_flat) / len(labels_flat)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "0DnMcDRcjKNL"
+ },
+ "source": [
+ "Helper function for formatting elapsed times.\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "BwYMBuRwjKNL"
+ },
+ "source": [
+ "import time\n",
+ "import datetime\n",
+ "\n",
+ "def format_time(elapsed):\n",
+ " '''\n",
+ " Takes a time in seconds and returns a string hh:mm:ss\n",
+ " '''\n",
+ " # Round to the nearest second.\n",
+ " elapsed_rounded = int(round((elapsed)))\n",
+ " \n",
+ " # Format as hh:mm:ss\n",
+ " return str(datetime.timedelta(seconds=elapsed_rounded))\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "uHdKDbX1jKNM",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 876
+ },
+ "outputId": "7f0b2f9c-6875-4645-ad77-c5961611d62a"
+ },
+ "source": [
+ "import random\n",
+ "\n",
+ "# This training code is based on the `run_glue.py` script here:\n",
+ "# https://github.com/huggingface/transformers/blob/5bfcd0485ece086ebcbed2d008813037968a9e58/examples/run_glue.py#L128\n",
+ "\n",
+ "\n",
+ "# Set the seed value all over the place to make this reproducible.\n",
+ "seed_val = 42\n",
+ "\n",
+ "random.seed(seed_val)\n",
+ "np.random.seed(seed_val)\n",
+ "torch.manual_seed(seed_val)\n",
+ "torch.cuda.manual_seed_all(seed_val)\n",
+ "\n",
+ "# Store the average loss after each epoch so we can plot them.\n",
+ "loss_values = []\n",
+ "\n",
+ "# For each epoch...\n",
+ "for epoch_i in range(0, epochs):\n",
+ " \n",
+ " # ========================================\n",
+ " # Training\n",
+ " # ========================================\n",
+ " \n",
+ " # Perform one full pass over the training set.\n",
+ "\n",
+ " print(\"\")\n",
+ " print('======== Epoch {:} / {:} ========'.format(epoch_i + 1, epochs))\n",
+ " print('Training...')\n",
+ "\n",
+ " # Measure how long the training epoch takes.\n",
+ " t0 = time.time()\n",
+ "\n",
+ " # Reset the total loss for this epoch.\n",
+ " total_loss = 0\n",
+ "\n",
+ " # Put the model into training mode. Don't be mislead--the call to \n",
+ " # `train` just changes the *mode*, it doesn't *perform* the training.\n",
+ " # `dropout` and `batchnorm` layers behave differently during training\n",
+ " # vs. test (source: https://stackoverflow.com/questions/51433378/what-does-model-train-do-in-pytorch)\n",
+ " model.train()\n",
+ "\n",
+ " # For each batch of training data...\n",
+ " for step, batch in enumerate(train_dataloader):\n",
+ "\n",
+ " # Progress update every 40 batches.\n",
+ " if step % 40 == 0 and not step == 0:\n",
+ " # Calculate elapsed time in minutes.\n",
+ " elapsed = format_time(time.time() - t0)\n",
+ " \n",
+ " # Report progress.\n",
+ " print(' Batch {:>5,} of {:>5,}. Elapsed: {:}.'.format(step, len(train_dataloader), elapsed))\n",
+ "\n",
+ " # Unpack this training batch from our dataloader. \n",
+ " #\n",
+ " # As we unpack the batch, we'll also copy each tensor to the GPU using the \n",
+ " # `to` method.\n",
+ " #\n",
+ " # `batch` contains three pytorch tensors:\n",
+ " # [0]: input ids \n",
+ " # [1]: attention masks\n",
+ " # [2]: labels \n",
+ " b_input_ids = batch[0].to(device)\n",
+ " b_input_mask = batch[1].to(device)\n",
+ " b_labels = batch[2].to(device)\n",
+ "\n",
+ " # Always clear any previously calculated gradients before performing a\n",
+ " # backward pass. PyTorch doesn't do this automatically because \n",
+ " # accumulating the gradients is \"convenient while training RNNs\". \n",
+ " # (source: https://stackoverflow.com/questions/48001598/why-do-we-need-to-call-zero-grad-in-pytorch)\n",
+ " model.zero_grad() \n",
+ "\n",
+ " # Perform a forward pass (evaluate the model on this training batch).\n",
+ " # This will return the loss (rather than the model output) because we\n",
+ " # have provided the `labels`.\n",
+ " # The documentation for this `model` function is here: \n",
+ " # https://huggingface.co/transformers/v2.2.0/model_doc/bert.html#transformers.BertForSequenceClassification\n",
+ " outputs = model(b_input_ids, \n",
+ " token_type_ids=None, \n",
+ " attention_mask=b_input_mask, \n",
+ " labels=b_labels)\n",
+ " \n",
+ " # The call to `model` always returns a tuple, so we need to pull the \n",
+ " # loss value out of the tuple.\n",
+ " loss = outputs[0]\n",
+ "\n",
+ " # Accumulate the training loss over all of the batches so that we can\n",
+ " # calculate the average loss at the end. `loss` is a Tensor containing a\n",
+ " # single value; the `.item()` function just returns the Python value \n",
+ " # from the tensor.\n",
+ " total_loss += loss.item()\n",
+ "\n",
+ " # Perform a backward pass to calculate the gradients.\n",
+ " loss.backward()\n",
+ "\n",
+ " # Clip the norm of the gradients to 1.0.\n",
+ " # This is to help prevent the \"exploding gradients\" problem.\n",
+ " torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0)\n",
+ "\n",
+ " # Update parameters and take a step using the computed gradient.\n",
+ " # The optimizer dictates the \"update rule\"--how the parameters are\n",
+ " # modified based on their gradients, the learning rate, etc.\n",
+ " optimizer.step()\n",
+ "\n",
+ " # Update the learning rate.\n",
+ " scheduler.step()\n",
+ "\n",
+ " # Calculate the average loss over the training data.\n",
+ " avg_train_loss = total_loss / len(train_dataloader) \n",
+ " \n",
+ " # Store the loss value for plotting the learning curve.\n",
+ " loss_values.append(avg_train_loss)\n",
+ "\n",
+ " print(\"\")\n",
+ " print(\" Average training loss: {0:.2f}\".format(avg_train_loss))\n",
+ " print(\" Training epcoh took: {:}\".format(format_time(time.time() - t0)))\n",
+ " \n",
+ " # ========================================\n",
+ " # Validation\n",
+ " # ========================================\n",
+ " # After the completion of each training epoch, measure our performance on\n",
+ " # our validation set.\n",
+ "\n",
+ " print(\"\")\n",
+ " print(\"Running Validation...\")\n",
+ "\n",
+ " t0 = time.time()\n",
+ "\n",
+ " # Put the model in evaluation mode--the dropout layers behave differently\n",
+ " # during evaluation.\n",
+ " model.eval()\n",
+ "\n",
+ " # Tracking variables \n",
+ " eval_loss, eval_accuracy = 0, 0\n",
+ " nb_eval_steps, nb_eval_examples = 0, 0\n",
+ "\n",
+ " # Evaluate data for one epoch\n",
+ " for batch in validation_dataloader:\n",
+ " \n",
+ " # Add batch to GPU\n",
+ " batch = tuple(t.to(device) for t in batch)\n",
+ " \n",
+ " # Unpack the inputs from our dataloader\n",
+ " b_input_ids, b_input_mask, b_labels = batch\n",
+ " \n",
+ " # Telling the model not to compute or store gradients, saving memory and\n",
+ " # speeding up validation\n",
+ " with torch.no_grad(): \n",
+ "\n",
+ " # Forward pass, calculate logit predictions.\n",
+ " # This will return the logits rather than the loss because we have\n",
+ " # not provided labels.\n",
+ " # token_type_ids is the same as the \"segment ids\", which \n",
+ " # differentiates sentence 1 and 2 in 2-sentence tasks.\n",
+ " # The documentation for this `model` function is here: \n",
+ " # https://huggingface.co/transformers/v2.2.0/model_doc/bert.html#transformers.BertForSequenceClassification\n",
+ " outputs = model(b_input_ids, \n",
+ " token_type_ids=None, \n",
+ " attention_mask=b_input_mask)\n",
+ " \n",
+ " # Get the \"logits\" output by the model. The \"logits\" are the output\n",
+ " # values prior to applying an activation function like the softmax.\n",
+ " logits = outputs[0]\n",
+ "\n",
+ " # Move logits and labels to CPU\n",
+ " logits = logits.detach().cpu().numpy()\n",
+ " label_ids = b_labels.to('cpu').numpy()\n",
+ " \n",
+ " # Calculate the accuracy for this batch of test sentences.\n",
+ " tmp_eval_accuracy = flat_accuracy(logits, label_ids)\n",
+ " \n",
+ " # Accumulate the total accuracy.\n",
+ " eval_accuracy += tmp_eval_accuracy\n",
+ "\n",
+ " # Track the number of batches\n",
+ " nb_eval_steps += 1\n",
+ "\n",
+ " # Report the final accuracy for this validation run.\n",
+ " print(\" Accuracy: {0:.2f}\".format(eval_accuracy/nb_eval_steps))\n",
+ " print(\" Validation took: {:}\".format(format_time(time.time() - t0)))\n",
+ "\n",
+ "print(\"\")\n",
+ "print(\"Training complete!\")"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "\n",
+ "======== Epoch 1 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:51.\n",
+ "\n",
+ " Average training loss: 0.15\n",
+ " Training epcoh took: 0:01:23\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:03\n",
+ "\n",
+ "======== Epoch 2 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:53.\n",
+ "\n",
+ " Average training loss: 0.03\n",
+ " Training epcoh took: 0:01:27\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:04\n",
+ "\n",
+ "======== Epoch 3 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:55.\n",
+ "\n",
+ " Average training loss: 0.01\n",
+ " Training epcoh took: 0:01:29\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:04\n",
+ "\n",
+ "======== Epoch 4 / 4 ========\n",
+ "Training...\n",
+ " Batch 40 of 66. Elapsed: 0:00:56.\n",
+ "\n",
+ " Average training loss: 0.00\n",
+ " Training epcoh took: 0:01:31\n",
+ "\n",
+ "Running Validation...\n",
+ " Accuracy: 1.00\n",
+ " Validation took: 0:00:04\n",
+ "\n",
+ "Training complete!\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "FKmiJDEijKNP"
+ },
+ "source": [
+ "Let's take a look at our training loss over all batches:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "AcPTD8-GjKNP",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 427
+ },
+ "outputId": "7e818d87-68de-4d41-938a-9b93c9b6b45a"
+ },
+ "source": [
+ "import matplotlib.pyplot as plt\n",
+ "% matplotlib inline\n",
+ "\n",
+ "import seaborn as sns\n",
+ "\n",
+ "# Use plot styling from seaborn.\n",
+ "sns.set(style='darkgrid')\n",
+ "\n",
+ "# Increase the plot size and font size.\n",
+ "sns.set(font_scale=1.5)\n",
+ "plt.rcParams[\"figure.figsize\"] = (12,6)\n",
+ "\n",
+ "# Plot the learning curve.\n",
+ "plt.plot(loss_values, 'b-o')\n",
+ "\n",
+ "# Label the plot.\n",
+ "plt.title(\"Training loss\")\n",
+ "plt.xlabel(\"Epoch\")\n",
+ "plt.ylabel(\"Loss\")\n",
+ "\n",
+ "plt.show()"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 864x432 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "_DNE4l0WjKNR"
+ },
+ "source": [
+ "# 4. Performance On Test Set"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "eYqb0aoXjKNS"
+ },
+ "source": [
+ "### 4.1. Data Preparation\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "GzlgmnhPjKNS",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "0650008c-15d7-4e39-bf96-460312da2ac5"
+ },
+ "source": [
+ "import pandas as pd\n",
+ "\n",
+ "# Load the dataset into a pandas dataframe.\n",
+ "df = pd.read_csv(\"/content/test_fin_07_06.csv\")\n",
+ "\n",
+ "#df = df.sample(3121)\n",
+ "# Report the number of sentences.\n",
+ "print('Number of test sentences: {:,}\\n'.format(df.shape[0]))\n",
+ "\n",
+ "# Create sentence and label lists\n",
+ "sentences = df.sentences.values\n",
+ "labels = df.labels.values\n",
+ "\n",
+ "\n"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Number of test sentences: 2,412\n",
+ "\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "PTdcVzRmjKQj",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 73
+ },
+ "outputId": "29f081ad-863d-4009-ee60-6d7bd89b92af"
+ },
+ "source": [
+ "df['labels'].value_counts()"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0 1871\n",
+ "1 541\n",
+ "Name: labels, dtype: int64"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 94
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "IBHRVcMSjKQl"
+ },
+ "source": [
+ "# Tokenize all of the sentences and map the tokens to thier word IDs.\n",
+ "input_ids = []\n",
+ "# For every sentence...\n",
+ "for sent in sentences:\n",
+ " # `encode` will:\n",
+ " # (1) Tokenize the sentence.\n",
+ " # (2) Prepend the `[CLS]` token to the start.\n",
+ " # (3) Append the `[SEP]` token to the end.\n",
+ " # (4) Map tokens to their IDs.\n",
+ " encoded_sent = tokenizer.encode(\n",
+ " sent, # Sentence to encode.\n",
+ " add_special_tokens = True, # Add '[CLS]' and '[SEP]'\n",
+ " )\n",
+ " \n",
+ " input_ids.append(encoded_sent)\n",
+ "\n",
+ "# Pad our input tokens\n",
+ "input_ids = pad_sequences(input_ids, maxlen=MAX_LEN, \n",
+ " dtype=\"long\", truncating=\"post\", padding=\"post\")\n",
+ "\n",
+ "# Create attention masks\n",
+ "attention_masks = []\n",
+ "\n",
+ "# Create a mask of 1s for each token followed by 0s for padding\n",
+ "for seq in input_ids:\n",
+ " seq_mask = [float(i>0) for i in seq]\n",
+ " attention_masks.append(seq_mask) \n",
+ "\n",
+ "# Convert to tensors.\n",
+ "prediction_inputs = torch.tensor(input_ids)\n",
+ "prediction_masks = torch.tensor(attention_masks)\n",
+ "prediction_labels = torch.tensor(labels)\n",
+ "\n",
+ "# Set the batch size. \n",
+ "batch_size = 32 \n",
+ "\n",
+ "# Create the DataLoader.\n",
+ "prediction_data = TensorDataset(prediction_inputs, prediction_masks, prediction_labels)\n",
+ "prediction_sampler = SequentialSampler(prediction_data)\n",
+ "prediction_dataloader = DataLoader(prediction_data, sampler=prediction_sampler, batch_size=batch_size)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Z5JKH6eojKQn"
+ },
+ "source": [
+ "## 5.2. Evaluate on Test Set\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "8BQjjXuNjKQo"
+ },
+ "source": [
+ "\n",
+ "With the test set prepared, we can apply our fine-tuned model to generate predictions on the test set."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "VAf5ckeLjKQo",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "58b3908d-a904-4182-ea25-ace37a103665"
+ },
+ "source": [
+ "# Prediction on test set\n",
+ "\n",
+ "print('Predicting labels for {:,} test sentences...'.format(len(prediction_inputs)))\n",
+ "\n",
+ "# Put model in evaluation mode\n",
+ "model.eval()\n",
+ "\n",
+ "# Tracking variables \n",
+ "predictions_test , true_labels = [], []\n",
+ "\n",
+ "# Predict \n",
+ "for batch in prediction_dataloader:\n",
+ "# Add batch to GPU\n",
+ " batch = tuple(t.to(device) for t in batch)\n",
+ " \n",
+ " # Unpack the inputs from our dataloader\n",
+ " b_input_ids, b_input_mask, b_labels = batch\n",
+ " \n",
+ " # Telling the model not to compute or store gradients, saving memory and \n",
+ " # speeding up prediction\n",
+ " with torch.no_grad():\n",
+ " # Forward pass, calculate logit predictions\n",
+ " outputs = model(b_input_ids, token_type_ids=None, \n",
+ " attention_mask=b_input_mask)\n",
+ "\n",
+ " logits = outputs[0]\n",
+ " #print(logits)\n",
+ "\n",
+ " # Move logits and labels to CPU\n",
+ " logits = logits.detach().cpu().numpy()\n",
+ " label_ids = b_labels.to('cpu').numpy()\n",
+ " #print(logits)\n",
+ " \n",
+ " # Store predictions and true labels\n",
+ " predictions_test.append(logits)\n",
+ " true_labels.append(label_ids)\n",
+ "\n",
+ "print(' DONE.')"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Predicting labels for 2,412 test sentences...\n",
+ " DONE.\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "WgWfACEJjKQq",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "fe9e69a1-894d-4c53-b447-d5a1ce04aecc"
+ },
+ "source": [
+ "predictions_test[0][0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([ 3.913034, -4.178549], dtype=float32)"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 97
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "shpd6OYJjKQs",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "d7455051-7f28-4cbc-9aa1-d97763926ba9"
+ },
+ "source": [
+ "print('Positive samples: %d of %d (%.2f%%)' % (df.labels.sum(), len(df.labels), (df.labels.sum() / len(df.labels) * 100.0)))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Positive samples: 541 of 2412 (22.43%)\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "pW-JXqXMjKQv",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "ac7a5b94-417d-4025-b617-a6be8f361237"
+ },
+ "source": [
+ "from sklearn.metrics import *\n",
+ "\n",
+ "matthews_set = []\n",
+ "pred_labels = []\n",
+ "f1_score_met = []\n",
+ "roc_auc_score_met = []\n",
+ "precision_score_met = []\n",
+ "recall_score_met = []\n",
+ "confusion_matrix_met =[]\n",
+ "precision_recall_curve_met =[]\n",
+ "accuracy_score_met = []\n",
+ "# Evaluate each test batch using Matthew's correlation coefficient\n",
+ "print('Calculating Matthews Corr. Coef. for each batch...')\n",
+ "\n",
+ "# For each input batch...\n",
+ "for i in range(len(true_labels)):\n",
+ " \n",
+ " # The predictions for this batch are a 2-column ndarray (one column for \"0\" \n",
+ " # and one column for \"1\"). Pick the label with the highest value and turn this\n",
+ " # in to a list of 0s and 1s.\n",
+ " pred_labels_i = np.argmax(predictions_test[i], axis=1).flatten()\n",
+ " pred_labels.append(pred_labels_i)\n",
+ "\n",
+ " accuracy_score_i = accuracy_score(pred_labels_i, true_labels[i])\n",
+ " accuracy_score_met.append(accuracy_score_i)\n",
+ "\n",
+ " f1_score_i = f1_score(pred_labels_i, true_labels[i], average=\"binary\")\n",
+ " f1_score_met.append(f1_score_i)\n",
+ "\n",
+ " confusion_matrix_i = confusion_matrix(pred_labels_i, true_labels[i])\n",
+ " confusion_matrix_met.append(confusion_matrix_i)\n",
+ "\n",
+ " precision_recall_curve_i = precision_recall_curve(pred_labels_i, true_labels[i])\n",
+ " precision_recall_curve_met.append(precision_recall_curve_i)\n",
+ "\n",
+ " precision_score_i = precision_score(pred_labels_i, true_labels[i], average=\"binary\")\n",
+ " precision_score_met.append(precision_score_i)\n",
+ "\n",
+ " roc_auc_score_i = roc_auc_score(pred_labels_i, true_labels[i])\n",
+ " roc_auc_score_met.append(roc_auc_score_i)\n",
+ "\n",
+ " recall_score_i = recall_score(pred_labels_i, true_labels[i], average=\"binary\")\n",
+ " recall_score_met.append(recall_score_i)\n",
+ "\n",
+ " # Calculate and store the coef for this batch depending on mathew\n",
+ " matthews = matthews_corrcoef(true_labels[i], pred_labels_i) \n",
+ " matthews_set.append(matthews)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Calculating Matthews Corr. Coef. for each batch...\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "s3X9y_uTjKQw"
+ },
+ "source": [
+ "#Calculate the number of Tp,Tn,Fp and Fn in the predictions\n",
+ "tns = 0\n",
+ "fps = 0\n",
+ "fns = 0\n",
+ "tps = 0\n",
+ "for i in range(len(predictions_test)):\n",
+ " pred_labels_i_fin = np.argmax(predictions_test[i], axis=1).flatten()\n",
+ " tn, fp, fn, tp = confusion_matrix(pred_labels_i_fin, true_labels[i]).ravel()\n",
+ " tns = tns + tn\n",
+ " fps = fps + fp\n",
+ " fns = fns + fn\n",
+ " tps = tps + tp"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "WtotHpQWjKQy",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "c9daf733-0494-468e-a1bf-85eb1e8ba8a3"
+ },
+ "source": [
+ "#Calculate the number of Tp,Tn,Fp and Fn in the predictions by confusion matrix\n",
+ "sum(confusion_matrix_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([[1801, 2],\n",
+ " [ 70, 539]])"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 101
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "PG6NwKSQjKQ0",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "2cab9729-9807-4052-c932-9acfa22faf21"
+ },
+ "source": [
+ "#Sample of true labels in the dataset\n",
+ "true_labels[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0,\n",
+ " 0, 1, 0, 0, 0, 0, 0, 0, 1, 0])"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 102
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "pvkZFlK8jKQ0",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "ffba2d86-95c2-411f-a6b5-8e584706dbb5"
+ },
+ "source": [
+ "#Sample of predicted labels in the predictions list\n",
+ "pred_labels[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "array([0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0,\n",
+ " 0, 1, 0, 0, 0, 0, 0, 0, 1, 0])"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 103
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "vFL5wsF1jKQ3"
+ },
+ "source": [
+ "import seaborn as sn\n",
+ "import matplotlib.pyplot as plt"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "oeqpj515jKQ5",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000
+ },
+ "outputId": "1e81adcf-85c6-4c96-dc70-5dbb41bb2138"
+ },
+ "source": [
+ "#Plot the confusion matrix for each batch\n",
+ "for i in range(len(confusion_matrix_met)):\n",
+ " df_cm = pd.DataFrame(confusion_matrix_met[i], index = [i for i in \"FT\"],\n",
+ " columns = [i for i in \"FT\"])\n",
+ " plt.figure(figsize = (3,2))\n",
+ " sn.heatmap(df_cm, annot=True)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\n",
+ " \"\"\"\n"
+ ],
+ "name": "stderr"
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQnklEQVR4nO3deVQUZ7oG8KeRTSCNoCAIKphBRFGISyaiiajIIBGBDIiaINwEjEZwLpg4ejUajSc6iiFG1CgaJYmiccRd48bEOOPCjUY9jKiRqAmytfFqo0Lv9w/GDp1quovtqy54f+fUOfZXRdWLpx++haJLotPpdCCEMGEldAGEdCQUOEIYosARwhAFjhCGKHCEMESBI4Qha6ELaEh1/yehS7AInXu8LHQJFkWtvGdyv7H3jU23Pm1VTotYVOAIaRaVQugKeKPAEdHTadRCl8AbBY6In5p6OEKYoR6OEJYocIQwRIsmhDBEPRwh7Oi0KqFL4I0CR8SPhpSEMERDSkIYosARwo6OhpSEMEQ9HCEMUeAIYUilFLoC3ihwRPyohyOEITX1cISwo6YezuLc+bkMh44V4mzRJfxSXgGFQoWeXp4IHzMSiZNi4dDZXn9s4IjxJs+VPn0a3k6a0tYlC0IikWB2egpSU9+AT29vyGQP8Pe/H8TiJavw9Gmt0OUZp9EIXQFvHSZwew8fR/6eQxg98o94NXw0rK2tUXTpKtZu+gLHCs9gx6Zs2NvZAQCWL3rP6DnWb/kKv9yrQOiIl1iWztTqrA8wOz0Fe/cdQXb2RgT080Na2psIDg5EeEQCLPKT8WnRxPKMCx2JlMQEPOfkqG9LiH0VvXv2wKa8nSg4eAxT4yYCAKL+NIbz9ZXVMtyrqMKAfn7w/4Mvs7pZ6t+/L9JmvYmCvYcxKWG6vv32nZ+x5pNlSEiIxs6d+wSssBEi6uE6zMfkBQb0NQjbMxFjXwEA/PjTXZNfv+/wCWi1Wvw5KqJN6rMEkxNiYGVlhU8/3WzQvnnLDjx58hSvT3lNoMrM0Ki5WxNcvXoVS5YsQWRkJIKDgxEaGoqMjAzcvct9T1y6dAlTpkxBUFAQRowYgWXLlqG2lv9Qm1fgcnNzUVpayv87EJGq6vsAgK6uXRo9RqfTYd+RE+jc2R6R40axKo25oUOCoNFoUPS/lw3aFQoFrlz5N4YODRaoMtN0KhVna4rNmzfjxIkTCAkJwYIFCzBp0iQUFRUhJibG4H1fUlKC5ORkKBQKzJs3D3Fxcdi1axcyMjJ4X4vXkHL16tXw8PDA888/DwB4+PAhxowZg40bN2LYsGFN+uYsiUajwWfb8mHdqRNeHTe60eMuXLyMsvJKxESOg5Mjt5dsLzx7dMf9+w+gVHLnRPfKKxESMgw2NjZQNfEN3ebULRtSJicnIysrC7a2tvq2yMhIREVFITc3FytWrAAAfPzxx+jSpQu+/PJLOP7nfeDt7Y2FCxfi3LlzGD58uNlrNWtIqdPp8PTpU6hFtBxrzN/WbMSV4hLMSkmEb2/vRo/bc/AYAOC1CeGsShOEQ+fOUCiML0DU1dXfIOzg0JllSfxoNNytCQYPHmwQNgDw8fGBn5+fvod7/Pgxzp49i5iYGH3YACA6OhoODg44evQor2t1mDnc763d9AV27DmI+OjxSJ2W0Ohxj+Q1OPXdWfj27onBQYEMK2TvaW0t7Oxsje6zt69fwbXIXw2oVNythXQ6He7fvw8XFxcAwI0bN6BWqxEYaPgesLW1RUBAAEpKSnidt8OsUja0bstX2JiXj5hXx2HRe+kmjz10rBBKpard924AUFFehf4BfWFra8sZVnr18IBM9qvlDScB6IwMKeVyOeRyOaddKpVCKpWaPeeBAwdQVVWln5/JZDIAgJubG+dYNzc3XL58mdNuDO/AVVRU4Pr16wCAmpoaAEBZWZm+7ff69evH99RMrdvyFTZ8vh3R48OwdN5/QyKRmDy+4NBxWFtbY+L4MEYVCuf7i1cQHh6KF4cF45//KtK329nZIShoAM6cOS9gdSYYGULm5eUhJyeH056Wlob0dNM/ZEtLS7F06VIMGTIE0dHRAIC6ujoA4Aw9gfr/n2f7zeEduOzsbGRnZxu0LVq0iHOcTqeDRCLh3cWytOHz7djw+XZERYzFh/+TASsr0yPq4pKbuHHrJ4SNGoGuLo2vYrYXX+8+gHl/Tcfs2SkGgUt5ayocHR2wY+deAaszQcldS0hKSkJsbCyn3VzvJpPJ8Pbbb8PZ2Rlr1qzRv0fs7evvRDK2oKRQKPT7zeEVuOXLl/M6mSXL33MQ67Z8Bc/u7nhpaDAOn/jWYH9Xly4IeXGwQVvBofrFkj9H/YlVmYIqLr6O9Ru2IW3Wm9j9dS6OHi3U32ly+vRZ5OdbaOCM9HB8h44N1dTUIDU1FTU1NcjPzzcYPj7797OhZUMymQzu7u68rsErcMZ+UohNcclNAEBFVTUWLFvN2T/0hYEGgatTKHD05Gl4dHfDiD8OYVan0DLnLMbdu2VISXkdkePH4v79B1i3bisWL1llmbd1wfgcrqkUCgVmzJiBO3fuYNu2bejTx/BxV3379oW1tTWKi4sRHv7bfF6pVKKkpARRUVG8riPRWdD/Ij0frh49H86QuefDPc6cyGlz+vgA7/NrNBqkpaXhu+++w/r16zFqlPGbG1JSUvDjjz/iyJEj+l8N7N69GwsXLsTWrVsREhJi9lodcpWStC86tbZFX79ixQoUFhZi9OjRePjwIfbv36/f5+joiLCw+gWzjIwMTJ48GYmJiYiPj0dlZSW2bt2KV155hVfYAOrhLBL1cIbM9XA1M7j3tz732Te8z5+YmIiioiKj+7y8vFBYWKh//f333yMrKwvXrl2Dk5MTIiMjkZmZCQcHB17XosBZIAqcIXOBk6dyf0cqzT3eVuW0CA0piei1dEjJEgWOiJ5OSYEjhB21xcyKzKLAEdHTUeAIYUerpMARwoxORH+WSYEjokeBI4Qhrcr0n1hZEgocET2tmgJHCDNaDQWOEGY0NKQkhB2tWjyfhUWBI6KnocARwo6G5nCEsKPVUA9HCDNqGlISwo5WS0PKZunmM07oEizC0G5+QpcgKhot9XCEMKOmORwh7Gh0NKQkhBkaUhLCkIp6OELY0YjoMYcUOCJ6GlAPRwgzKgocIeyozTxU05JQ4IjotfxhVeyIZ7ZJSCNUEglna6rq6mpkZWUhMTERL7zwAvz9/XHhwgWjx546dQqxsbEYOHAgQkNDkZOTA7Wa3ycZUeCI6KklEs7WVLdv30Zubi6qqqrg7+/f6HGnT5/GrFmz4OzsjPfffx9hYWFYt24d76cE05CSiF5rfMLCgAEDcP78ebi4uODkyZOYNWuW0eNWrlyJ/v37Y8uWLejUqROA+mfIbdq0CYmJifDx8TF5HerhiOhpJNytqZycnODi4mLymFu3buHWrVtISEjQhw0Apk6dCq1Wi+PHzT8ii3o4InrGZk9yuRxyuZzTLpVKIZVKm3Wda9euAQACAwMN2rt37w4PDw/9flMocET0jA0p8/LykJOTw2lPS0tDenp6s64jk8kAAG5ubpx9bm5uqK6uNnsOChwRPWNDyP9KSkJsbCynvbm9GwDU1dUBAGxtbTn77OzsUFtba/YcFDgiesaGlC0ZOjbG3t4eAKBUKjn7FAqFfr8ptGhCRE8l4W5t4dlQ8tnQsiGZTAZ3d3ez56DAEdHTQMfZ2kJAQAAAoLi42KC9qqoKlZWV+v2mUOCI6GmMbG3Bz88Pffr0wa5du6DR/HaV/Px8WFlZITw83Ow5OvwcLnPODAQFD0BwcCB8fHvh7t0yDBowSuiyLIJdZzvsKNwKr949sHvrXqxesEbokoxSSlqnR1u/fj0AoLS0FACwf/9+XLx4EVKpFG+88QYAYO7cuZg5cybeeustREZG4ubNm9i+fTsSEhLg6+tr9homA1deXg5XV1dek0GxWrzkPTz49f9w5cq/4ezcupNssZv+3pvo0rWL0GWY1Vo92po1hj9Q9uzZAwDw8vLSB2706NHIyclBTk4OPvzwQ7i6umLmzJl45513eF3DZODGjh2LlStXIioqqjn1i0JQYCju3PkFAHCu6CgcHR0Ersgy+A/0Q0JKHNYt+wx/+cD4bU6WorXmbDdu3OB1XFhYGMLCwpp1DZNzOJ1OPA8rb65nYSO/sbKywvxV7+L8P4rw7ZEzQpdjlgo6zmapOvwcjnBNnh6P3n/ohfkpi4UuhZe2WpVsC2ZXKSUi+mta0nKePT2Q+m4yPs/+AhVllUKXw0u76uE++ugjZGdn8zqZRCLByZMnW1wUEc5f/5aJe3crsGPj10KXwpvaggP2e2YD5+npCQ8PDxa1EIFFvDYOL74yFDNf+ws0avF8cIGYhpRmA5ecnNyuVylJPRtbG8z+4B2cPXUBv1Y/gLePFwDAzaMbAMDpOUd4+3jh4YNHeCx/LGSpHCqdVugSeKNFEwIAsLO3g2s3F4wcNxwjxw3n7B8fF47xceH4dOkG7PhslwAVNq5d9XCkY6h9Wov5qdxVSZeuzpi7IhPnCi/gQP4R3CopFaA60yhwIpIwOQY9e9UPn7p1c4WtjQ3enVv/i95ffr6HXTv3CVkeMxq1Bv84fJrT7uldP38vu1tudL8lUKGdDCmvX7/Oqg7BJCbF4+WXXzJoe39RJgDgzJnzHSZwYqahOZx4TBj/utAlWLSKskq81CNU6DJMoiElIQzRKiUhDGnayxyOEDGgORwhDKkpcISwo6YhJSHs0JCSEIbUOvHcaE2BI6JHPRwhDFEPRwhD1MMRwhAFjhCG1Fp+z9e2BBQ4InrUwxHCEAWOEIbUWvGsUtLTc4joaXRaztZUSqUSq1atwsiRIzFo0CBMmjQJ586da/VaKXBE9DRaLWdrqnnz5iEvLw8TJ07EggULYGVlhdTUVPzwww+tWqtEZ0EPEHB2el7oEixCgLSn0CVYlPPl35rc7+TAfUzU46e3eZ//6tWriI+Px/z585GcnAyg/hHCEyZMgLu7O7Zv396Eak2jHo6IXkuHlN988w1sbGwQHx+vb7Ozs0NcXBwuXryI6urqVquVFk2I6GmNBEwul0Mul3PapVIppFLD5wCWlJTA19cXjo6OBu2DBg2CTqdDSUkJr+d382FRgXv02PI+85BYPkUd95Fja9euRU5ODqc9LS0N6enpBm0ymQzdu3fnHOvm5gYA1MMRYk5SUhJiY2M57b/v3QCgrq4ONjY2nHY7OzsA9fO51kKBI+2SsaFjY+zt7aFSqTjtz4L2LHitgRZNSIfn5uZmdNgok8kAoNXmbwAFjhD069cPt2/fxpMnTwzar1y5ot/fWihwpMOLiIiASqXC7t279W1KpRIFBQUYPHiw0QWV5qI5HOnwgoKCEBERgaysLMhkMvTq1Qt79+5FeXk5li9f3qrXsqg7TQgRikKhwCeffIKDBw/i0aNH8Pf3R2ZmJkJCQlr1OhQ4QhiiORwhDFHgCGGIFk0AFBQUYP78+Ub3zZkzB9OnT2dcEXv+/v68jjt16hS8vb3buJr2iwLXQEZGBjw9PQ3a+vfvL1A1bK1cudLgdV5eHsrLyzk/iFxdXVmW1e5Q4BoYNWoUAgIChC5DENHR0Qavjx07hocPH3LaScvQHI4QhqiHa0Aul+PBgwf61xKJBC4uLgJWRNobClwD06ZNM3jt4ODQ6p9pQTo2ClwDS5YsQa9evfSvO3XqJGA1pD2iwDUQFBTUYRdNCBu0aEIIQxQ4QhiiwBHCEAWOEIYocIQwRH8PRwhD1MMRwhAFjhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQxQ4QhiiwBHCEAWOEIb+H4Ntxrr/3JJGAAAAAElFTkSuQmCC\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAASUElEQVR4nO3deVxU5f4H8M8MwrCJgIIgqGgpQibgUleuWSpyEVOkxHGJ4hpYBri2YItLVppgRI0LohYlIrkQGrmkeLmVC4mBcQW7ENpFBIa4iLIMs/3+4MdcpjMyw3bOHPi+X6/zejnPOcz5ji8+PM955sw8ArVarQYhhBVCrgsgpC+hwBHCIgocISyiwBHCIgocISyiwBHCon5cF9CWvPo3rkswChZDnuC6BKOiaL7d7n5dvzemg0b2VDldYlSBI6RT5DKuKzAYBY7wnlqp4LoEg1HgCP8pqIcjhDXUwxHCJgocISyiSRNCWEQ9HCHsUavkXJdgMAoc4T8aUhLCIhpSEsIiChwh7FHTkJIQFlEPRwiLKHCEsEjezHUFBqPAEf6jHo4QFimohyOEPQrq4YzOzd/L8M3pLFzIuYr/lN+BTCbHUBdn+E+fgtAFwbC0MNc6Pq+gEHu/SEPhr8WovXsPDoPs8fgEb4SHLsBQF2eOXkXPEwgEWBEdjoiI5+A23BVSaQ2OHDmBDZti0dDQyHV5uimVXFdgsD4TuPTMM0g9+g2mTXkcs/2noV+/fsi5eg2f7vkCp7O+x8E98TAXiQAAP1y6glde24ChLs5Y9Oxc2A6wQUnpLRw5fhLf/eMHpH+5C4MdBnH8inrG9riNWBEdjvSvv0V8fCI8xoxCVNRSeHuPhX+AGEb5zfg0aWJ8Zj41BeGhYvS3ttK0iYNnY/jQIdiTfAjHTpzG4vlzAQBfpKVDKBTiwO7tsLMdoDn+oRHDsfHDBJzJ+h6h4mDWX0NP8/QcjajIpTiWnokF4mWa9tKbvyPh4/cgFgfh0KGvOazwAXjUw/WZr8kb6zFaK2ytAmZMBQD8+7dbmrb6+gaIzExh099a61jHQfYAAIs/DT97i4XieRAKhfjkk71a7Xv3HUR9fQOWLHqGo8r0UCqYm5EyKHBJSUkoKSnp6Vo4UVlVDQAYaG+rafN9fALqGxrx5nvbUfTv31AprcaPl3MRK9mLkW5DMWvGk1yV26MmTvCCUqlEzk95Wu0ymQz5+f/CxIneHFXWPrVcztiMlUFDyu3bt8PJyQkPPfQQAKC2thbTp09HYmIiJk2a1KMF9iSlUondn6ein4kJZs+cpmmPCF2Amv/WIj3zDDLPnNe0PzF5EmI3vQErK0suyu1xzkMGo7q6Bs3NzGui2+UV8PWdBFNTU8iN7RdawZ8hZaeu4dRqNRoaGqDg0XSsLh8mJCK/oBArXwrDiOGumnah0ASODgPxl4k+mDHVFwNsrPHzL9dx8MhxvLZ+Kz75cANM+/W+y19LCwvIZLonIJqaWm4QtrS0wN27Rha4Ll7DXbt2Denp6bh8+TLKy8tha2sLHx8frFq1CsOHD9c69urVq4iNjcX169dhbW2NWbNmYe3atbCwsDDoXL3vt8ZAn+75AgePnkBI0CxEPC/W2vfW+9uRX1CIrw/s1sxc+j35VwxzGYLNcRJkfHsW8+cGcFF2j2pobISjjutcADA3b/l/MMq3BrrY4+7duxdXr15FQEAA3N3dIZVKkZKSgnnz5uHIkSOakV1hYSHCwsLw8MMPIyYmBhUVFdi/fz/Kysqwe/dug87VJwO3Y98BJCanYt7smVj/WrTWvjsVVcg8cx6Ln52jCVurv01/ApvjJLiS90uvDNyd8kp4eoyGmZkZY1jpMsQJUukfxjecBKDu4pAyLCwMcXFxMDMz07QFBgZizpw5SEpKwtatWwEAH330EWxtbfHll1/CyqrlD5OrqyvefvttXLx4EZMnT9Z7LoNnKe/cuYOioiIUFRWhuLgYAFBWVqZp+/NmrHbsO4Bd+1MQNMsP78asgkAg0NpfWf0HAECpUjF+VvH/Qxclj6ahO+JKbj5MTEzw2CTtyRGRSAQvr0eQm5vPUWV6KJXMrQPGjx+vFTYAcHNzw6hRozSThffv38eFCxcwb948TdgAICgoCJaWljh58qRB5zK4h4uPj0d8fLxW2/r16xnHqdVqCAQCFBYWGvrUrNm1PwW79qdgTsAMbH5zNYRC5t+bEcNcYWIiRNY/L2LlS2Fabw1kfPsdgJa3GHqjrw4fR8wb0VixIhw//JijaQ9/cTGsrCxx8FA6h9W1o7n75xLUajWqq6sxZswYAMCNGzegUCgwduxYrePMzMzg4eFh8O+7QYHbsmVLB8s1PqlHT2DHvgNwHuyIv0z0RuZ3/9DaP9DOFr6PjccAm/54bsE8JKceQ8jfo/DsnAAMsOmPn3+5jswz5zHUxRnPzvkbNy+ihxUUFGHnrs8RFbkUh79KwsmTWZo7TbKzLyA11UgDp6NHq6urQ11dHaPdxsYGNjY2ep/y+PHjqKysxOrVqwEAUqkUAODg4MA41sHBAXl5eYx2XQwKXHAw/++qKCj8FQBwp7IKb723nbF/os+j8H1sPADg1chwjBjmiqMnTiPpyzTIm+VwdBgIcfBsvLJ0CaytdE8s9AZr1m7ArVtlCA9fgsBZM1BdXYMdOz7Dhk2xxnlbF3RfwyUnJ0MikTDao6KiEB0dzWhvq6SkBO+++y4mTJiAoKAgAEBTUxMAMIaeQMuQu3W/PgK1Ef0v0vpwLWh9OG361oe7v2Yuo0218UCnejipVIpFixZBpVIhLS1N06OdOnUKK1euxKFDh+Dj46P1MytXrkReXh6ys7P1vpY+OUtJehe1gjnBZejQsa179+4hIiIC9+7dQ2pqqtbwsfXfrUPLtqRSKRwdHQ06R5+5l5L0Ys0K5tZBMpkML7/8Mm7evInExESMHKm9guro0aPRr18/FBQUaJ+6uRmFhYXw8PAw6DwUOMJ7aqWKsXWEUqnEqlWrkJeXh4SEBHh7M+8Z7d+/PyZPnoyMjAzU19dr2jMyMtDQ0ICAAMPel6UhJeE9XUPKjti6dSuysrIwbdo01NbWIiMjQ7PPysoKfn5+AIDVq1dj4cKFCA0NRUhICCoqKvDZZ59h6tSp8PX1NehcNGlihGjSRJu+SZPaJdMZbbYpWQY/f2hoKHJycnTuc3FxQVbW/57rypUriIuL09xLGRgYiDVr1sDS0rAb2ilwRogCp01v4MTTGG22aed1HMk9GlIS3lMrjKbP0IsCR3hP1UyBI4Q1ah59LJMCR3iPAkcIi1Rygf6DjAQFjvCeSkGBI4Q1KiUFjhDWKGlISQh7VAr+3BJMgSO8p6TAEcIeJV3DEcIelZJ6OEJYo6AhJSHsUaloSNkp9LGUFuMHPcx1CbyiVFEPRwhrFHQNRwh7lGoaUhLCGhpSEsIiOfVwhLBHyaNve6TAEd5Tgno4Qlgjp8ARwh6FgAJHCGv4tB4tBY7wnpxHPRx/pncIeQCFQMDYOqqqqgpxcXEIDQ2Fj48P3N3dcfnyZZ3Hnjt3DsHBwXj00Ufx1FNPQSKRQKEw7KvDKHCE9+QC5tZRpaWlSEpKQmVlJdzd3R94XHZ2NiIjIzFgwAC888478PPzw44dOwxelpuGlIT3uuPzp4888gguXboEOzs7nD17FpGRkTqP27ZtGzw9PbFv3z6YmJgAaFlhZ8+ePQgNDYWbm1u756EejvCeQsfWUdbW1rCzs2v3mOLiYhQXF0MsFmvCBgCLFy+GSqXCmTNn9J6HejjCe7qGkHV1dZ1a47s9169fBwCMHTtWq33w4MFwcnLS7G8PBY7wnq4hZXJyMiQSCaM9KioK0dHRnTpP6/rebdf+buXg4ICqqiq9z0GBI7ynawgZ8cILCA4OZrR3tncDgKamJgCAmZkZY59IJEJjY6Pe56DAEd7TNaTsytDxQczNzQEAzc3NjH0ymUyzvz00aUJ4Twk1Y+sJrUPJ1qFlW1KpFI6OjnqfgwJHeE+pY+sJHh4eAICCggKt9srKSlRUVGj2t4cCB0AgEGDliggU/JKN+3UlKC35CbEfroelpQXXpXFKZCFC+sVU5JRn49X3V3JdzgM1C9SMrSeMGjUKI0eORFpaGpTK/8U6NTUVQqEQ/v7+ep+j3Wu48vJy2NvbGzQ25bPtcRuxIjoc6V9/i/j4RHiMGYWoqKXw9h4L/wAx1Gr+LGnbnV56bSlsB9pyXYZe3dWj7dy5EwBQUlICAMjIyEBubi5sbGzw3HPPAQBef/11LF++HC+++CICAwPx66+/IiUlBWKxGCNGjNB7jnYDN2PGDGzbtg1z5szp6msxWp6eoxEVuRTH0jOxQLxM015683ckfPwexOIgHDr0NYcVcsP90VFYGD4fkvcSsWqj7rsujEV3XbMlJCRoPT569CgAwMXFRRO4adOmQSKRQCKRYPPmzbC3t8fy5cvxyiuvGHSOdgPXF/6yLxTPg1AoxCef7NVq37vvID54/00sWfRMnwucUCjEm7Gv4dL5HJz/9p9GHzh5NwXuxo0bBh3n5+cHPz+/Tp2jz78tMHGCF5RKJXJ+ytNql8lkyM//FyZO9OaoMu4sWhYCt4eHISZ8PdelGKSnZiV7gt5JEwGPPmvUGc5DBqO6ukbneyu3yyvg4DAQpqamHFTGjSFDnbDs1b9jX3wy7pRVcF2OQeRQMzZjpbeH++CDDxAfH2/QkwkEApw9e7bLRbHJ0sICMhkzbADQ1CRrOcbSAnfvytksizMxH67F7Vt3kJL4FdelGExhxAH7M72Bc3Z2hpOTExu1cKKhsRGO1lY695mbi1qOadB/y05vEPDMTDw2dSJeemYFlAr+fHEBn4aUegMXFhbWq2cp75RXwtNjNMzMzBjDSpchTpBK/4Bc3vt7N1MzU6zaGIkL5y7hj6oauLq5AAAcnAYBAKz7W8PVzQW1NXdxv+4+l6UyyNUqrkswWJ+fNLmSmw9//6fw2CRv/PBjjqZdJBLBy+sRfP/9JQ6rY4/IXAT7QXaYMtMXU2b6MvYHzvdH4Hx/JLy7Eym70zio8MF6VQ/X2311+Dhi3ojGihXhWoELf3ExrKwscfBQOofVsaexoRExEcxZSbuBtnhj6xpcyLqM46mZKC4s4aC69lHgeKSgoAg7d32OqMilOPxVEk6ezNLcaZKdfQGpqX0jcEqFElmZ2Yx2Z9eW6/eyW7d17jcGcvSSIWVRURFbdXBqzdoNuHWrDOHhSxA4awaqq2uwY8dn2LAptk+8+c93Sh5dwwnURvQb1c/MhesSjAKtgKotp7z9nvXpYbMZbd/8ntlT5XRJnx9SEv6jWUpCWKTsLddwhPABn67hKHCE9xQUOELYo6AhJSHsoSElISxSqPlzozUFjvAe9XCEsIh6OEJYRD0cISyiwBHCIoWqMyvCcYMCR3iPejhCWESBI4RFChV/ZilpMQ/Ce0q1irF1VHNzM2JjYzFlyhSMGzcOCxYswMWLF7u9Vgoc4T2lSsXYOiomJgbJycmYO3cu3nrrLQiFQkRERODnn3/u1lrpE99GiD7xrU3fJ76tLZmr1txvKDX4+a9du4aQkBCsW7cOYWFhAFq+6v7pp5+Go6MjUlJSOlRve6iHI7zX1SHlqVOnYGpqipCQEE2bSCTC/PnzkZubi6qqqm6rlSZNCO+pdASsrq4OdXV1jHZda38XFhZixIgRsLLS/gbucePGQa1Wo7Cw0KDlhA1hVIFTNN/mugTCQ7Km/zDaPv30U0gkEkZ7VFQUoqOjtdqkUikGDx7MOLZ1TW/q4QjR44UXXkBwcDCj/c+9GwA0NTXpXCFJJGpZW0Imk3VbXRQ40ivpGjo+iLm5uc71I1qD1hq87kCTJqTPc3Bw0DlslEqlANBt128ABY4QjBkzBqWlpaivr9dqz8/P1+zvLhQ40ucFBARALpfj8OHDmrbm5mYcO3YM48eP1zmh0ll0DUf6PC8vLwQEBCAuLg5SqRTDhg1Deno6ysvLsWXLlm49l1HdaUIIV2QyGT7++GOcOHECd+/ehbu7O9asWQNfX+ZaeV1BgSOERXQNRwiLKHCEsIgmTQAcO3YM69at07lv7dq1WLZsGcsVsc/d3d2g486dOwdXV9cerqb3osC1sXr1ajg7O2u1eXp6clQNu7Zt26b1ODk5GeXl5Yw/RPb29myW1etQ4Np48skn4eHhwXUZnAgKCtJ6fPr0adTW1jLaSdfQNRwhLKIero26ujrU1NRoHgsEAtjZ2XFYEeltKHBtPP/881qPLS0tu/07LUjfRoFrY9OmTRg2bJjmsYmJCYfVkN6IAteGl5dXn500IeygSRNCWESBI4RFFDhCWESBI4RFFDhCWESfhyOERdTDEcIiChwhLKLAEcIiChwhLKLAEcIiChwhLKLAEcIiChwhLKLAEcIiChwhLPo/QMahHFxWojkAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQsklEQVR4nO3de1RTV74H8G94v4xAGwSx+GiRlwoqauU6PimDehG4FXCsFKaV1lbwgrReO23ttdOlVlG0glbxxVgL1gfiY6y9iOPSUdErVYYKemV8DMMr1GoUJO/7hzVDPJEECDs54fdZ66zV7H3I+dWVb/Y+Oyc5ArVarQYhhAkrUxdASG9CgSOEIQocIQxR4AhhiAJHCEMUOEIYsjF1Ae3Jm/9u6hLMgmP/35i6BLOikP2zw35drxvbF4f0VDndYlaBI6RL5FJTV2AwChzhPbVSYeoSDEaBI/ynoBGOEGZohCOEJQocIQzRogkhDNEIRwg7apXc1CUYjAJH+I+mlIQwRFNKQhiiwBHCjrqbU8qKigoUFRWhrKwMdXV1cHV1xciRI5Geno6BAwdq7VteXo41a9bg2rVrcHFxwfTp05GZmQlHR0eDjkWBI/zXzRFu27ZtKC8vR2RkJPz8/CAWi7Fnzx7ExMRg//79ePnllwEAVVVVSE5OxiuvvIKlS5eioaEBO3bsQG1tLb7++muDjkWBI/zXzcAlJycjKysLdnZ2mrYZM2YgKioKeXl5WLVqFQBg3bp1cHV1xe7du+Hs7AwAGDBgAD755BOcP38e48eP13ss+j4c4T+5jLt1wqhRo7TCBgCDBg2Cr68vampqAACPHj3CuXPnEBMTowkbAERHR8PJyQnHjx836FgUOMJ/SgV36ya1Wo3m5ma4ubkBAK5fvw6FQoFhw4Zp7WdnZ4eAgABUVVUZ9Lw0pST8p+COaBKJBBKJhNMuFAohFAr1PuXhw4fR2NiIjIwMAIBYLAYAiEQizr4ikQhXrlwxqFQKHOE/BXdEy8/PR05ODqc9NTUVaWlpHT5dTU0NPv/8c4wePRrR0dEAgLa2NgDgTD0BwN7eXtOvT68J3O27tTh6ohTnLpbjH3X1kErleMnbCxFTJyAxPhZOjg7P/dvCoqP4IisXAHDmWCHcXPuyKps5gUCARWnzkZIyD4MGDoBYfA/79x/BZ8vXoLX1sanL002p5DQlJSUhNjaW065vdBOLxXj33XfRt29fbNiwAVZWT866HByevD5kMu5oKpVKNf369JrAFR37AQUHjmLKhHGYGTEFNjY2uFhegY1b/4QTpWfw7dZsONjbc/6uSfwz1m/eCSdHR7Q+NtMXnBGtzfpvLEqbj6JDf0Z29hYE+PsiNfUthIQMQ0RkAszyl/F1LJIYOnVs7+HDh0hJScHDhw9RUFCgNX18+t9Pp5bticVieHh4GHSMXhO41yZPwPzEBPRx+dcKU0LsTAx8qT+25hfi4JETmDt7FufvvliXi5e8vfDy4IE4eqKUZcnMBQYORerCt3Cw6BjiE97RtN+6fRcb1n+BhIRoFBYeMmGFz6FjhOssqVSKBQsW4Pbt29i1axeGDNH+EaKhQ4fCxsYGlZWViIiI0LTLZDJUVVUhKirKoOP0mlXKYQFDtcL2VOS0iQCA//v7HU5fyem/4i9ny7DswzRYW1n+P9WchBhYWVnhq6+2abVv2/4tWlpa8cbv/sNElenRzVVKpVKJ9PR0XLlyBRs2bEBISAhnnz59+mD8+PEoLi5GS0uLpr24uBitra2IjIw06FgGjXB5eXmYOnWq5hN3S9LY1AwAeMHdVav9UUsLVqzbjLjo6Rge6IfCg0dNUR5ToaODoVQqcfGS9oqbVCrF1as/ITSU+0I0B2p5976es2rVKpSWlmLKlCm4f/8+iouLNX3Ozs4IDw8HAGRkZGDOnDlITExEXFwcGhoasHPnTkycOBFhYWEGHcugwK1duxaenp6awN2/fx9Tp07Fli1bMGbMmM7+/5kNpVKJr3cVwMbaGjNfm6LVt27TDqjUKqQv+L2JqmPPq38/NDff07kw8M+6BoSFjYGtrS3k3XyBG52ie1PK6upqAMCpU6dw6tQprT5vb29N4IKCgrBz505kZWVh5cqVcHFxQXx8PBYvXmzwsbp0DqdWq9Ha2gqFjuVYPvlywxZcrazCf76bjMEDB2jayyt+wr7i4/jysyU6p6GWysnREVKp7qs02tqeXCDs5OSIBw/MLHDdPIfbvXu3wfuGhoaisLCwy8fqNYsmz9q49U/49sARxEVPR8qbCZp2uVyO5V9+hVdDQzDjtcmmK9AEWh8/hsdz3mAcHJ6s4JrlRwPmNuJ2oFcGLnf7N9iSX4CYma9h2YfaH4IWHDiKW3dr8WFaCu7W1mnaW359odXWN+BRSyte8vZiWjML9XWNCAwYCjs7O8600ru/J8Tin81vOglA3c0pJUsGB66+vl4z13348CEAoLa2VtP2LH9/fyOUZ3y527/B5h17ED09HJ8vTYdAINDqr2tohEqlwoLMT3X+/e/mp8PR0QGXSopYlMvU/16+ioiIyRg7JgRn/3pR025vb4/g4CCcOXPBhNV1wAgfC7BicOCys7ORnZ2t1bZs2TLOfmq1GgKBwOCLOVnavGMPNu/Yg6jIafjjHzI0VxG0FzMzAqOCgzjtBQeO4tKPFfjjHzIg7OPColzmvtt3GEv/Kw2LFs3XCtz8t+fC2dkJ3xaa6ZuMjD9rCQYFbuXKlT1dR48rOHAEudu/gVc/D7waGoJj//MXrf4X3FwRNnYU/H2HwN+Xe+eV07++ACf/2ziLvbSrsrIamzbvQurCt7DvuzwcP16qudLk9OlzKCgw08BZ2gin65o0vqmsugEAqG9swsdfrOX0h44cjrCxo1iXZXYWZ36GO3dqMX/+G5gxfRqam+8hN3cnPlu+xjwv6wK/zuEEajP6V6T7wz1B94fTpu/+cI8Wcy/Jc1l3uKfK6ZZeuUpJLItaoTJ1CQajwBH+s7RFE0LMmVpJIxwhzNCUkhCG1DIKHCHsKMxmoV0vChzhPTUFjhB2VDIKHCHMqPnzqQAFjvAfBY4QhlRygf6dzAQFjvCeSkGBI4QZlZICRwgzSppSEsKOSsGfH+mlwBHeU1LgCGFHSedwhLCjUtIIRwgzCppSEsKOSkVTyi7xGBShf6deYBH9iFCnKFU0whHCjILO4QhhR6mmKSUhzNCUkhCG5Dwa4fjz1kDIcyhhxdk6q6mpCVlZWUhMTMTIkSPh5+eHsrIynfuePHkSsbGxGD58OCZPnoycnByDb05KgSO8p4SAs3XWrVu3kJeXh8bGRvj5+T13v9OnT2PhwoXo27cvPv30U4SHhyM3N9fgG97QlJLwnrwLAXtWUFAQLly4ADc3N5SUlGDhwoU691u9ejUCAwOxfft2WFtbAwCcnZ2xdetWJCYmYtCgQR0eh0Y4wnsKgYCzdZaLiwvc3Nw63OfmzZu4efMmEhISNGEDgLlz50KlUuGHH37Qexwa4Qjv6bpZlUQigUQi4bQLhUIIhcIuHefatWsAgGHDhmm19+vXD56enpr+jlDgCO/JdYxo+fn5yMnJ4bSnpqYiLS2N024IsVgMABCJRJw+kUiEpqYmvc9BgSO8p2sKmZSUpPNGol0d3QCgra0NAGBnZ8fps7e3x+PHj/U+BwWO8J6uX1joztTxeRwcHAAAMpmM0yeVSjX9HaFFE8J7SgF36wlPp5JPp5bticVieHh46H0OChzhPYWOrScEBAQAACorK7XaGxsb0dDQoOnvCAWO8J5cwN16gq+vL4YMGYK9e/dCqfzX2mhBQQGsrKwQEaH/62V0Dkd4z1hTyE2bNgEAampqAADFxcW4fPkyhEIh5s2bBwBYsmQJ3nvvPbz99tuYMWMGbty4gT179iAhIQGDBw/WewyBWq02m1uPuLm8YuoSzMLvXww1dQlmZd3twg771/rM47Rl3v2m08d53iVd3t7eKC0t1TwuKSlBTk4Oampq4O7ujtdffx3vv/8+bGz0j180whHeM9YU8vr16wbtFx4ejvDw8C4dgwJHeE8Js5mk6UWBI7yn69Iuc9XrA5eRuQAjQoIQEhKEQYN9cPdOLYKDJpu6LJP4bfps/DZ99nP7lXIFPvTlni+ZmkxgISNcXV0d3N3dDfoEna+WLf8A937+BVev/oS+fY17ZQLfVHx/Ec23GzjtXv4+mLpgFn46WW6CqvSzmBFu2rRpWL16NaKioljVw1zIsCm4c/sfAIBzF/8MZ2cnE1dkOvXVd1FffZfTHrdiPgCgbO8p1iUZxGLO4czoE4Me8zRsRDc7R3uE/HsYfqn7GdWnr5i6HJ3kPAocXWlCOhQ881U4Cp1waf9pqFXm+cJWQs3ZzJXeRRNBF749SyzHuPgpUKlUuPideU4nAX6NcHoDt2LFCmRnZxv0ZAKBACUlJd0uipgH0RAvDBnrjxtn/4Z7tdwr5M2FwpIC5+XlBU9PTxa1EDMzLn4KAODC3lI9e5qWOU8hn6U3cMnJyRa9Skl0s7K2QujrE/HongR/O3HJ1OV0SK5WmboEg9GiCdEpKHw0hCJXlB86C6Wsp75hZhwWtWhCeqexv04nzfWzt/bMOWDP6vWBS5gTgwE+/QEAL7zoDjtbW2QueR8AUHu3DnsLD5myPJMQerjBf1Iw7ly5ifrr5v85pRz8mVJ2GLjq6mpWdZjMvKQ4TPjNOK22T5YtBgCcPVPWKwM3ZvYkWNtYo6zQvBdLnlLy6Byu149wUdPfMHUJZufkpkM4uYk/bzQ0pSSEIT6tUlLgCO8pLeUcjhA+oHM4QhhSUOAIYUdBU0pC2KEpJSEMKdT8+ZEFChzhPRrhCGGIRjhCGKIRjhCGKHCEMKRQmff39dqjwBHeoxGOEIYocIQwpFDxZ5WSftOE8J5SreJsnSWTybBmzRpMmDABI0aMQHx8PM6fP2/0WilwhPeUKhVn66ylS5ciPz8fs2bNwscffwwrKyukpKTgxx9/NGqtFDjCe3KVkrN1RkVFBY4dO4YPPvgAS5YsQUJCAvLz8+Hl5YWsrCyj1kqBI7zX3Snl999/D1tbW8TFxWna7O3tMXv2bFy+fBlNTU1Gq5UWTQjvqXQETCKRQCKRcNqFQiGEQu37AFZVVWHw4MFwdnbWah8xYgTUajWqqqrg4eFhlFrNKnC/PLpp6hIID0nbuD/lt3HjRuTk5HDaU1NTkZaWptUmFovRr18/zr4ikQgAaIQjRJ+kpCTExsZy2p8d3QCgra0Ntra2nHZ7e3sAgFQqNVpdFDhikXRNHZ/HwcEBcrmc0/40aE+DZwy0aEJ6PZFIpHPaKBY/uUWXsc7fAAocIfD398etW7fQ0tKi1X716lVNv7FQ4EivFxkZCblcjn379mnaZDIZDh48iFGjRulcUOkqOocjvV5wcDAiIyORlZUFsVgMHx8fFBUVoa6uDitXrjTqsQRqtZo/P8xOSA+RSqVYv349jhw5ggcPHsDPzw+LFy9GWFiYUY9DgSOEITqHI4QhChwhDNGiCYCDBw/io48+0tmXmZmJd955h3FF7Pn5+Rm038mTJzFgwIAersZyUeDaycjIgJeXl1ZbYGCgiapha/Xq1VqP8/PzUVdXx3kjcnd3Z1mWxaHAtTNp0iQEBASYugyTiI6O1np84sQJ3L9/n9NOuofO4QhhiEa4diQSCe7du6d5LBAI4ObmZsKKiKWhwLXz5ptvaj12cnIy+m9akN6NAtfO8uXL4ePjo3lsbW1twmqIJaLAtRMcHNxrF00IG7RoQghDFDhCGKLAEcIQBY4QhihwhDBE34cjhCEa4QhhiAJHCEMUOEIYosARwhAFjhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQ/8Pn2jIm3jXxGQAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARo0lEQVR4nO3de1hUdf4H8PdwvzUKhYCat37IxQuk0ha6eSMWcRHIy6SGmoJdAHex1rWflj/6+aSrGJloKZqypWgkqHhNZFdrtdgwcVnBktBSBIZ8cEhgLmfm9wc/ZxnPMDMIfM8c5vN6nvM8zvccz/noM2++lznMkeh0Oh0IIUzYCV0AIbaEAkcIQxQ4QhiiwBHCEAWOEIYocIQw5CB0Ae2pG34UugSr4Nr/t0KXYFU0qlsm9xt73zg+NqynyukSqwocIQ9FrRS6AotR4Ijo6TiN0CVYjAJHxE9DPRwhzFAPRwhLFDhCGKJFE0IYoh6OEHZ0WrXQJViMAkfEj4aUhDBEQ0pCGKLAEcKOjoaUhDBEPRwhDFHgCGFIrRK6AotR4Ij4UQ9HCEMa6uEIYUdDPZzVuf7TTRw9VYzzJRfxc81tKJVqPD7AD5FTJiBhTjzcXF30x44cP83kuVKXLsDLC+f2dMmCkEgkWJaaiKSkFzFk8EDI5Xfw+eeFWJO+Ec3NLUKXZxzHCV2BxWwmcAXHvkDuwaOYPOE3mB45GQ4ODii5eBlbdvwVp4q/xL4dmXBxdgYArHv7T0bPsW3Xp/j51m1MGv80y9KZ2pTxP1iWmoiCQ8eRmbkdQYH+SElZjNDQkYiMksEqvxmfFk2sz3OTJiAxQYZHPNz1bbL46Rj8eH/syNmP/MJTmDdrBgAg5ndTeH+/tl6OW7frMCLQHwH/NZRZ3SwFBw9HSvJi5BccwxzZUn179fWfsPn9tZDJYrF//yEBK+yAiHo4m/mavJFBww3Cdl/U1GcBAD/8eMPk3z907DS0Wi1mxkT1SH3W4AVZHOzs7PDBBzsN2nfu2od795oxf+7zAlVmBqfhb51w+fJlpKenIzo6GqGhoZg0aRLS0tJw4wb/PXHx4kXMnTsXISEhGD9+PNauXYuWFsuH2hYFLjs7G1VVVZb/C0Skrr4BAPCoV98Oj9HpdDh0/DRcXV0Q/dxEVqUxN25sCDiOQ8k/Lxm0K5VKlJX9G+PGhQpUmWk6tZq3dcbOnTtx+vRphIeHY9WqVZgzZw5KSkoQFxdn8L6vqKjAokWLoFQqsXLlSsyaNQsHDhxAWlqaxdeyaEi5adMm+Pr64oknngAANDY2YsqUKdi+fTvCwsI69Y+zJhzH4aM9uXCwt8f05yZ3eNw3pZdws6YWcdHPwcOd30v2Fn79fdDQcAcqFX9OdKumFuHhYXB0dIS6k2/oHqfp2pBy0aJFyMjIgJOTk74tOjoaMTExyM7Oxvr16wEA7733Hvr27YtPPvkE7v//Phg4cCBWr16NCxcu4JlnnjF7rYcaUup0OjQ3N0MjouVYY/6yeTvKyiuQnJiAoYMHdnjcwcJTAIDnfx/JqjRBuLm6Qqk0vgDR2tp2g7CbmyvLkizDcfytE8aMGWMQNgAYMmQI/P399T3cr7/+ivPnzyMuLk4fNgCIjY2Fm5sbTpw4YdG1bGYO96AtO/6KfQcLMTt2GpIWyDo87q6iCWfOncfQwY9jTMhIhhWy19zSAmdnJ6P7XFzaVnCt8qMBtZq/dZFOp0NDQwM8PT0BAFevXoVGo8HIkYbvAScnJwQFBaGiosKi89rMKmV7W3d9iu05uYib/hze/lOqyWOPniqGSqXu9b0bANyuqUNw0HA4OTnxhpUD+vtCLv/F+oaTAHRGhpQKhQIKhYLXLpVKIZVKzZ7zyJEjqKur08/P5HI5AMDb25t3rLe3Ny5dusRrN8biwN2+fRuVlZUAgKamJgDAzZs39W0PCgwMtPTUTG3d9Sk+/HgvYqdF4J2Vf4REIjF5fP7RL+Dg4IAZ0yIYVSicb0vLEBk5CU+FheKrf5To252dnRESMgJffvm1gNWZYGQImZOTg6ysLF57SkoKUlNN/5CtqqrCO++8g7FjxyI2NhYA0NraCgC8oSfQ9v9zf785FgcuMzMTmZmZBm1vv/027zidTgeJRGJxF8vShx/vxYcf70VM1FT873+nwc7O9Ii6vOJ7XL32IyImjsejnh2vYvYWn+Udwco/p2LZskSDwCUumQd3dzfs218gYHUmqPhrCQsXLkR8fDyv3VzvJpfL8fLLL6NPnz7YvHmz/j3i4tJ2J5KxBSWlUqnfb45FgVu3bp1FJ7NmuQcLsXXXp/Dz6Yenx4Xi2Om/G+x/1LMvwp8aY9CWf7RtsWRmzO9YlSmo8vJKbPtwD1KSFyPvs2ycOFGsv9Pk7NnzyM210sAZ6eEsHTq219TUhKSkJDQ1NSE3N9dg+Hj/z/eHlu3J5XL069fPomtYFDhjPynEprziewDA7bp6rFq7ibd/3JOjDALXqlTiRNFZ+Pp4Y/xvxjKrU2jLX1+DGzduIjFxPqKnTUVDwx1s3boba9I3WudtXTA+h+sspVKJV155BdevX8eePXswbJjh466GDx8OBwcHlJeXIzLyP/N5lUqFiooKxMTEWHQdic6K/hfp+XBt6Plwhsw9H+7X5TN4bR7vHbH4/BzHISUlBefOncO2bdswcaLxmxsSExPxww8/4Pjx4/qPBvLy8rB69Wrs3r0b4eHhZq9lk6uUpHfRabRd+vvr169HcXExJk+ejMbGRhw+fFi/z93dHRERbQtmaWlpeOGFF5CQkIDZs2ejtrYWu3fvxrPPPmtR2ADq4awS9XCGzPVwTa/w72995KOTFp8/ISEBJSUlRvcNGDAAxcXF+tfffvstMjIycOXKFXh4eCA6OhrLly+Hm5ubRdeiwFkhCpwhc4FTJPE/I5Vmf9FT5XQJDSmJ6HV1SMkSBY6Ink5FgSOEHY3VzIrMosAR0dNR4AhhR6uiwBHCjE5Ev5ZJgSOiR4EjhCGt2vSvWFkTChwRPa2GAkcIM1qOAkcIMxwNKQlhR6sRz3dhUeCI6HEUOELY4WgORwg7Wo56OEKY0dCQkhB2tFoaUj4U+k3nNlN9RgtdgqhwWurhCGFGQ3M4QtjhdDSkJIQZGlISwpCaejhC2OFE9JhDChwRPQ7UwxHCjJoCRwg7GjMP1bQmFDgiel1/WBU74pltEtIBtUTC2zqrvr4eGRkZSEhIwJNPPomAgAB88803Ro89c+YM4uPjMWrUKEyaNAlZWVnQaCz7JiMKHBE9jUTC2zqruroa2dnZqKurQ0BAQIfHnT17FsnJyejTpw/eeustREREYOvWrRY/JZiGlET0uuMbFkaMGIGvv/4anp6eKCoqQnJystHjNmzYgODgYOzatQv29vYA2p4ht2PHDiQkJGDIkCEmr0M9HBE9TsLfOsvDwwOenp4mj7l27RquXbsGmUymDxsAzJs3D1qtFl98Yf4RWdTDEdEzNntSKBRQKBS8dqlUCqlU+lDXuXLlCgBg5MiRBu0+Pj7w9fXV7zeFAkdEz9iQMicnB1lZWbz2lJQUpKamPtR15HI5AMDb25u3z9vbG/X19WbPQYEjomdsCPnSwoWIj4/ntT9s7wYAra2tAAAnJyfePmdnZ7S0tJg9BwWOiJ6xIWVXho4dcXFxAQCoVCrePqVSqd9vCi2aENFTS/hbT7g/lLw/tGxPLpejX79+Zs9BgSOix0HH23pCUFAQAKC8vNygva6uDrW1tfr9plDgiOhxRrae4O/vj2HDhuHAgQPguP9cJTc3F3Z2doiMjDR7DprDAZBIJFiWmoikpBcxZPBAyOV38PnnhViTvhHNzeYnwr3JqZ9PGG1vudeCuMDnGVdjGZWke3q0bdu2AQCqqqoAAIcPH0ZpaSmkUilefPFFAMCKFSvw6quvYsmSJYiOjsb333+PvXv3QiaTYejQoWavIdHpdB1WW1NTAy8vL4smg93BwWkAk+s86L1N6ViWmoiCQ8dx8uTfEBToj+Tkl/DVVyWIjJLBxH9RjxDyW7tO/XwC//rmXzi+1zB4nIbD2cJzgtVkyptD5vHa1l3f1+nrdHRL14ABA1BcXKx/XVRUhKysLFRVVcHLywszZ87Ea6+9BgcH8/2XySOmTp2KDRs2ICYmppOli0dw8HCkJC9GfsExzJEt1bdXX/8Jm99fC5ksFvv3HxKwQvZu/1SL4oK/CV2Gxbprznb16lWLjouIiEBERMRDXcPkHI71T3YhvCCLg52dHT74YKdB+85d+3DvXjPmz7XOYVRPc3B0gIsbm5FNV6mh423WyubncOPGhoDjOJT885JBu1KpRFnZvzFuXKhAlQnnt9ETMDV+Cuwd7NHY0IizheewZ2MOmpuahS7NqJ5alewJZgMnEdFv0z4Mv/4+aGi4Y/TDzFs1tQgPD4OjoyPUarUA1bFX+V0lvjz2FW5dr4G7hxvCpoQh9qUZGPX0KKTFLUdrc6vQJfJYc4/2ILOBe/fdd5GZmWnRySQSCYqKirpcFEturq5QKvlhA4DWVmXbMW6uuHvXNgL3hxlpBq+LDp5BdUU1XvrzIsQviUPulv0CVdYxTW8KnJ+fH3x9fVnUIojmlhb083A3us/FxbntGBv7aOBBeR99jvl/nI+npoRZZeB61ZBy0aJFvXqV8nZNHYKDhsPJyYk3rBzQ3xdy+S82M5zsCKfhcKfuF0i9uvfexO6i1mmFLsFiNn+nybelZbC3t8dTYYaLI87OzggJGYHS0jKBKrMejs6OeMzvMTQ2NApdilGsbu3qDjYfuM/yjkCr1WLZskSD9sQl8+Du7oZ9+wsEqoy9R/o+YrR94RsL4ODogK9PG/9SHaGJKXA2/7FAeXkltn24BynJi5H3WTZOnChGUKA/UlIW4+zZ88jNtZ3AzfvDXAQ+GYjLF8pQf0sOV3dXhE0eh9Dxoai4WInDe44IXaJRaohnSGkycJWVlazqENTy19fgxo2bSEycj+hpU9HQcAdbt+7GmvSNNvHh/32XL1zGIP9BiJgVAWlfKbRaLW5V38Luv+zBwex8qJXWOZflRDSHM3kvJWtC3UtpbegJqIbM3Uv5+0HTeW1HfzrWU+V0ic0PKYn4iWmVkgJHRI/rLXM4QsRATHM4ChwRPQ0FjhB2NDSkJIQdGlISwpBGJ54nxFHgiOhRD0cIQ9TDEcIQ9XCEMESBI4Qhjday52tbAwocET3q4QhhiAJHCEMarXhWKW3+KxaI+HE6LW/rLJVKhY0bN2LChAkYPXo05syZgwsXLnR7rRQ4InqcVsvbOmvlypXIycnBjBkzsGrVKtjZ2SEpKQnfffddt9ZKgSOip9ZyvK0zLl++jGPHjuGNN97AihUrIJPJkJOTAz8/P2RkZHRrrRQ4InpdHVKePHkSjo6OmD17tr7N2dkZs2bNQmlpKerr67utVlo0IaKnNRIwhUIBhULBa5dKpZBKDb/QtqKiAkOHDoW7u+E3cI8ePRo6nQ4VFRUWPb/bElYVOI3qltAlEBFStv7Ma9uyZQuysrJ47SkpKUhNTTVok8vl8PHx4R3r7e0NANTDEWLOwoULER8fz2t/sHcDgNbWVjg6OvLanZ3bni2hVCq7rS4KHOmVjA0dO+Li4mL0+RH3g3Y/eN2BFk2IzfP29jY6bJTL5QDQbfM3gAJHCAIDA1FdXY179+4ZtJeVlen3dxcKHLF5UVFRUKvVyMvL07epVCrk5+djzJgxRhdUHhbN4YjNCwkJQVRUFDIyMiCXyzFo0CAUFBSgpqYG69at69ZrWdWzBQgRilKpxPvvv4/CwkLcvXsXAQEBWL58OcLDw7v1OhQ4QhiiORwhDFHgCGGIFk0A5Ofn48033zS67/XXX8fSpUsZV8ReQECARcedOXMGAwcO7OFqei8KXDtpaWnw8/MzaAsODhaoGrY2bNhg8DonJwc1NTW8H0ReXl4sy+p1KHDtTJw4EUFBQUKXIYjY2FiD16dOnUJjYyOvnXQNzeEIYYh6uHYUCgXu3Lmjfy2RSODp6SlgRaS3ocC1s2DBAoPXbm5u3f6dFsS2UeDaSU9Px6BBg/Sv7e3tBayG9EYUuHZCQkJsdtGEsEGLJoQwRIEjhCEKHCEMUeAIYYgCRwhD9PtwhDBEPRwhDFHgCGGIAkcIQxQ4QhiiwBHCEAWOEIYocIQwRIEjhCEKHCEMUeAIYej/AN8wY3BHv/8AAAAAAElFTkSuQmCC\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQsklEQVR4nO3de1RTV74H8G94v4xAGwSx+GiRlwoqauU6PimDehG4FXCsFKaV1lbwgrReO23ttdOlVlG0glbxxVgL1gfiY6y9iOPSUdErVYYKemV8DMMr1GoUJO/7hzVDPJEECDs54fdZ66zV7H3I+dWVb/Y+Oyc5ArVarQYhhAkrUxdASG9CgSOEIQocIQxR4AhhiAJHCEMUOEIYsjF1Ae3Jm/9u6hLMgmP/35i6BLOikP2zw35drxvbF4f0VDndYlaBI6RL5FJTV2AwChzhPbVSYeoSDEaBI/ynoBGOEGZohCOEJQocIQzRogkhDNEIRwg7apXc1CUYjAJH+I+mlIQwRFNKQhiiwBHCjrqbU8qKigoUFRWhrKwMdXV1cHV1xciRI5Geno6BAwdq7VteXo41a9bg2rVrcHFxwfTp05GZmQlHR0eDjkWBI/zXzRFu27ZtKC8vR2RkJPz8/CAWi7Fnzx7ExMRg//79ePnllwEAVVVVSE5OxiuvvIKlS5eioaEBO3bsQG1tLb7++muDjkWBI/zXzcAlJycjKysLdnZ2mrYZM2YgKioKeXl5WLVqFQBg3bp1cHV1xe7du+Hs7AwAGDBgAD755BOcP38e48eP13ss+j4c4T+5jLt1wqhRo7TCBgCDBg2Cr68vampqAACPHj3CuXPnEBMTowkbAERHR8PJyQnHjx836FgUOMJ/SgV36ya1Wo3m5ma4ubkBAK5fvw6FQoFhw4Zp7WdnZ4eAgABUVVUZ9Lw0pST8p+COaBKJBBKJhNMuFAohFAr1PuXhw4fR2NiIjIwMAIBYLAYAiEQizr4ikQhXrlwxqFQKHOE/BXdEy8/PR05ODqc9NTUVaWlpHT5dTU0NPv/8c4wePRrR0dEAgLa2NgDgTD0BwN7eXtOvT68J3O27tTh6ohTnLpbjH3X1kErleMnbCxFTJyAxPhZOjg7P/dvCoqP4IisXAHDmWCHcXPuyKps5gUCARWnzkZIyD4MGDoBYfA/79x/BZ8vXoLX1sanL002p5DQlJSUhNjaW065vdBOLxXj33XfRt29fbNiwAVZWT866HByevD5kMu5oKpVKNf369JrAFR37AQUHjmLKhHGYGTEFNjY2uFhegY1b/4QTpWfw7dZsONjbc/6uSfwz1m/eCSdHR7Q+NtMXnBGtzfpvLEqbj6JDf0Z29hYE+PsiNfUthIQMQ0RkAszyl/F1LJIYOnVs7+HDh0hJScHDhw9RUFCgNX18+t9Pp5bticVieHh4GHSMXhO41yZPwPzEBPRx+dcKU0LsTAx8qT+25hfi4JETmDt7FufvvliXi5e8vfDy4IE4eqKUZcnMBQYORerCt3Cw6BjiE97RtN+6fRcb1n+BhIRoFBYeMmGFz6FjhOssqVSKBQsW4Pbt29i1axeGDNH+EaKhQ4fCxsYGlZWViIiI0LTLZDJUVVUhKirKoOP0mlXKYQFDtcL2VOS0iQCA//v7HU5fyem/4i9ny7DswzRYW1n+P9WchBhYWVnhq6+2abVv2/4tWlpa8cbv/sNElenRzVVKpVKJ9PR0XLlyBRs2bEBISAhnnz59+mD8+PEoLi5GS0uLpr24uBitra2IjIw06FgGjXB5eXmYOnWq5hN3S9LY1AwAeMHdVav9UUsLVqzbjLjo6Rge6IfCg0dNUR5ToaODoVQqcfGS9oqbVCrF1as/ITSU+0I0B2p5976es2rVKpSWlmLKlCm4f/8+iouLNX3Ozs4IDw8HAGRkZGDOnDlITExEXFwcGhoasHPnTkycOBFhYWEGHcugwK1duxaenp6awN2/fx9Tp07Fli1bMGbMmM7+/5kNpVKJr3cVwMbaGjNfm6LVt27TDqjUKqQv+L2JqmPPq38/NDff07kw8M+6BoSFjYGtrS3k3XyBG52ie1PK6upqAMCpU6dw6tQprT5vb29N4IKCgrBz505kZWVh5cqVcHFxQXx8PBYvXmzwsbp0DqdWq9Ha2gqFjuVYPvlywxZcrazCf76bjMEDB2jayyt+wr7i4/jysyU6p6GWysnREVKp7qs02tqeXCDs5OSIBw/MLHDdPIfbvXu3wfuGhoaisLCwy8fqNYsmz9q49U/49sARxEVPR8qbCZp2uVyO5V9+hVdDQzDjtcmmK9AEWh8/hsdz3mAcHJ6s4JrlRwPmNuJ2oFcGLnf7N9iSX4CYma9h2YfaH4IWHDiKW3dr8WFaCu7W1mnaW359odXWN+BRSyte8vZiWjML9XWNCAwYCjs7O8600ru/J8Tin81vOglA3c0pJUsGB66+vl4z13348CEAoLa2VtP2LH9/fyOUZ3y527/B5h17ED09HJ8vTYdAINDqr2tohEqlwoLMT3X+/e/mp8PR0QGXSopYlMvU/16+ioiIyRg7JgRn/3pR025vb4/g4CCcOXPBhNV1wAgfC7BicOCys7ORnZ2t1bZs2TLOfmq1GgKBwOCLOVnavGMPNu/Yg6jIafjjHzI0VxG0FzMzAqOCgzjtBQeO4tKPFfjjHzIg7OPColzmvtt3GEv/Kw2LFs3XCtz8t+fC2dkJ3xaa6ZuMjD9rCQYFbuXKlT1dR48rOHAEudu/gVc/D7waGoJj//MXrf4X3FwRNnYU/H2HwN+Xe+eV07++ACf/2ziLvbSrsrIamzbvQurCt7DvuzwcP16qudLk9OlzKCgw08BZ2gin65o0vqmsugEAqG9swsdfrOX0h44cjrCxo1iXZXYWZ36GO3dqMX/+G5gxfRqam+8hN3cnPlu+xjwv6wK/zuEEajP6V6T7wz1B94fTpu/+cI8Wcy/Jc1l3uKfK6ZZeuUpJLItaoTJ1CQajwBH+s7RFE0LMmVpJIxwhzNCUkhCG1DIKHCHsKMxmoV0vChzhPTUFjhB2VDIKHCHMqPnzqQAFjvAfBY4QhlRygf6dzAQFjvCeSkGBI4QZlZICRwgzSppSEsKOSsGfH+mlwBHeU1LgCGFHSedwhLCjUtIIRwgzCppSEsKOSkVTyi7xGBShf6deYBH9iFCnKFU0whHCjILO4QhhR6mmKSUhzNCUkhCG5Dwa4fjz1kDIcyhhxdk6q6mpCVlZWUhMTMTIkSPh5+eHsrIynfuePHkSsbGxGD58OCZPnoycnByDb05KgSO8p4SAs3XWrVu3kJeXh8bGRvj5+T13v9OnT2PhwoXo27cvPv30U4SHhyM3N9fgG97QlJLwnrwLAXtWUFAQLly4ADc3N5SUlGDhwoU691u9ejUCAwOxfft2WFtbAwCcnZ2xdetWJCYmYtCgQR0eh0Y4wnsKgYCzdZaLiwvc3Nw63OfmzZu4efMmEhISNGEDgLlz50KlUuGHH37Qexwa4Qjv6bpZlUQigUQi4bQLhUIIhcIuHefatWsAgGHDhmm19+vXD56enpr+jlDgCO/JdYxo+fn5yMnJ4bSnpqYiLS2N024IsVgMABCJRJw+kUiEpqYmvc9BgSO8p2sKmZSUpPNGol0d3QCgra0NAGBnZ8fps7e3x+PHj/U+BwWO8J6uX1joztTxeRwcHAAAMpmM0yeVSjX9HaFFE8J7SgF36wlPp5JPp5bticVieHh46H0OChzhPYWOrScEBAQAACorK7XaGxsb0dDQoOnvCAWO8J5cwN16gq+vL4YMGYK9e/dCqfzX2mhBQQGsrKwQEaH/62V0Dkd4z1hTyE2bNgEAampqAADFxcW4fPkyhEIh5s2bBwBYsmQJ3nvvPbz99tuYMWMGbty4gT179iAhIQGDBw/WewyBWq02m1uPuLm8YuoSzMLvXww1dQlmZd3twg771/rM47Rl3v2m08d53iVd3t7eKC0t1TwuKSlBTk4Oampq4O7ujtdffx3vv/8+bGz0j180whHeM9YU8vr16wbtFx4ejvDw8C4dgwJHeE8Js5mk6UWBI7yn69Iuc9XrA5eRuQAjQoIQEhKEQYN9cPdOLYKDJpu6LJP4bfps/DZ99nP7lXIFPvTlni+ZmkxgISNcXV0d3N3dDfoEna+WLf8A937+BVev/oS+fY17ZQLfVHx/Ec23GzjtXv4+mLpgFn46WW6CqvSzmBFu2rRpWL16NaKioljVw1zIsCm4c/sfAIBzF/8MZ2cnE1dkOvXVd1FffZfTHrdiPgCgbO8p1iUZxGLO4czoE4Me8zRsRDc7R3uE/HsYfqn7GdWnr5i6HJ3kPAocXWlCOhQ881U4Cp1waf9pqFXm+cJWQs3ZzJXeRRNBF749SyzHuPgpUKlUuPideU4nAX6NcHoDt2LFCmRnZxv0ZAKBACUlJd0uipgH0RAvDBnrjxtn/4Z7tdwr5M2FwpIC5+XlBU9PTxa1EDMzLn4KAODC3lI9e5qWOU8hn6U3cMnJyRa9Skl0s7K2QujrE/HongR/O3HJ1OV0SK5WmboEg9GiCdEpKHw0hCJXlB86C6Wsp75hZhwWtWhCeqexv04nzfWzt/bMOWDP6vWBS5gTgwE+/QEAL7zoDjtbW2QueR8AUHu3DnsLD5myPJMQerjBf1Iw7ly5ifrr5v85pRz8mVJ2GLjq6mpWdZjMvKQ4TPjNOK22T5YtBgCcPVPWKwM3ZvYkWNtYo6zQvBdLnlLy6Byu149wUdPfMHUJZufkpkM4uYk/bzQ0pSSEIT6tUlLgCO8pLeUcjhA+oHM4QhhSUOAIYUdBU0pC2KEpJSEMKdT8+ZEFChzhPRrhCGGIRjhCGKIRjhCGKHCEMKRQmff39dqjwBHeoxGOEIYocIQwpFDxZ5WSftOE8J5SreJsnSWTybBmzRpMmDABI0aMQHx8PM6fP2/0WilwhPeUKhVn66ylS5ciPz8fs2bNwscffwwrKyukpKTgxx9/NGqtFDjCe3KVkrN1RkVFBY4dO4YPPvgAS5YsQUJCAvLz8+Hl5YWsrCyj1kqBI7zX3Snl999/D1tbW8TFxWna7O3tMXv2bFy+fBlNTU1Gq5UWTQjvqXQETCKRQCKRcNqFQiGEQu37AFZVVWHw4MFwdnbWah8xYgTUajWqqqrg4eFhlFrNKnC/PLpp6hIID0nbuD/lt3HjRuTk5HDaU1NTkZaWptUmFovRr18/zr4ikQgAaIQjRJ+kpCTExsZy2p8d3QCgra0Ntra2nHZ7e3sAgFQqNVpdFDhikXRNHZ/HwcEBcrmc0/40aE+DZwy0aEJ6PZFIpHPaKBY/uUWXsc7fAAocIfD398etW7fQ0tKi1X716lVNv7FQ4EivFxkZCblcjn379mnaZDIZDh48iFGjRulcUOkqOocjvV5wcDAiIyORlZUFsVgMHx8fFBUVoa6uDitXrjTqsQRqtZo/P8xOSA+RSqVYv349jhw5ggcPHsDPzw+LFy9GWFiYUY9DgSOEITqHI4QhChwhDNGiCYCDBw/io48+0tmXmZmJd955h3FF7Pn5+Rm038mTJzFgwIAersZyUeDaycjIgJeXl1ZbYGCgiapha/Xq1VqP8/PzUVdXx3kjcnd3Z1mWxaHAtTNp0iQEBASYugyTiI6O1np84sQJ3L9/n9NOuofO4QhhiEa4diQSCe7du6d5LBAI4ObmZsKKiKWhwLXz5ptvaj12cnIy+m9akN6NAtfO8uXL4ePjo3lsbW1twmqIJaLAtRMcHNxrF00IG7RoQghDFDhCGKLAEcIQBY4QhihwhDBE34cjhCEa4QhhiAJHCEMUOEIYosARwhAFjhCGKHCEMESBI4QhChwhDFHgCGGIAkcIQ/8Pn2jIm3jXxGQAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACWCAYAAAC1meaLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARxklEQVR4nO3de1RU9d7H8fcgONwcAQUkvJcCYqJmnaPL002OoWcpekrplLdKuwj0eCmP1ake7Zw0xdTCLpJ6OGl21dDM6vGSp9K0NDUTLSkviOAQ0aDIMLfnD5ezHPfIjAJ7ZsP3tdZey/ntzeyva/Hh99u339Y5HA4HQghVBPi6ACGaEwmcECqSwAmhIgmcECqSwAmhIgmcECoK9HUBF7OU/+zrEvxCyDV/8nUJfsVae7LO9e5+b4Ladm2scurFrwInxFWxmH1dgdckcELzHDarr0vwmgROaJ9VejghVFPfHm7//v2sXbuWnTt3UlJSQkREBH369GHKlCl06tTJZds9e/Ywf/58Dh48SHh4OEOGDGH69OmEhIR4tS8JnNC+egbujTfeYM+ePaSlpZGQkIDRaGTVqlWMGDGC999/n2uvvRaAwsJCJkyYwHXXXcfMmTMpLS1l+fLlFBcX89prr3m1Lwmc0L56njSZMGECOTk5tGzZ0tk2dOhQhg0bRl5eHnPnzgXgxRdfJCIigjfffJOwsDAA2rdvzz/+8Q927NhB//79Pe5LrsMJ7bNZlcsV6Nu3r0vYADp37ky3bt0oKioC4MyZM2zfvp0RI0Y4wwaQnp5OaGgoGzdu9Gpf0sMJzXPYLYo2k8mEyWRStBsMBgwGg+fvdDgoLy8nMTERgMOHD2O1WunZs6fLdi1btiQpKYnCwkKvapXACe1zM6TMz88nNzdX0Z6VlUV2drbHr1y3bh1lZWVMnToVAKPRCEB0dLRi2+joaPbu3etVqRI4oX1uhpDjx49n5MiRinZvereioiJmz57NDTfcQHp6OgA1NTUAiqEngF6vd673RAIntM9N4LwdOl7KaDTy0EMP0bp1axYvXkxAwPnTHMHBwQDU1tYqfsZsNjvXeyKBE5rnaKBbu6qqqpg0aRJVVVWsXr3aZfh44d8XhpYXMxqNxMTEeLUPOUsptK+eZynhfC/18MMPc/ToUV5//XW6dnW9+bl79+4EBgZy4MABl/ba2loKCwtJSkryaj8SOKF99QyczWZjypQp7N27l8WLF9O7d2/FNq1ataJ///4UFBRw9uxZZ3tBQQHV1dWkpaV5tS8ZUgrtsyiPq67E3Llz2bJlC7fddhuVlZUUFBQ414WFhZGamgrA1KlTufvuuxk7diyjRo2itLSUFStWcPPNNzNgwACv9iWBE9pXz1u7Dh06BMDWrVvZunWry7r4+Hhn4JKTk1mxYgU5OTnMmTOH8PBwRo8ezbRp07zel86f5qWUB1DPkwdQXXl6APXcmucVbSF/fbKxyqkX6eGE9lnleTi/c/R4MR99uoXtu/ZwouQUZrOFDvFxDL59IGNHjyQ05Px1FIfDwUefbWXbVzv54dBPGMsriIgwkHhdVx4cfze9khN9/D9pXDqdjkezJzJp0hg6d2qP0VjB+++v59lZ86muPufr8tyz2XxdgdeazZBy4avLWf3BR9w28A/0Sk4kMDCQXXv28+mW/9L9ui68tXQhwXo9ZnMtN9yeTmK3rtw84Cbi49pR/msF7364gdPlFTz/9GMMu+P2RqsTfDukfHHBLB7NnsjaDz/mk0+2kpTYjczM+/jyy10MTsvAF78uHoeU/3lC0RYybk5jlVMvzSZwBwp/pFOHeFqFh7m0v7Q0n6X5b/Pk1Ee4567hWK02vvv+B27s08tlu/KK3xgx5mECAgL4fN0q5x0IjcFXgevRozt792zmw4KNjM540NmeOfk+Fi/6J2PGZfL22x+qXpfHwK2YoWgLuW9eY5VTL83mOlzPpO6KsAGkDboZgJ9+PgZAYGALRdgA2kZF0q/39VT8VknFb5WNW6yP3J0xgoCAAF566Q2X9jeWvcXZs9Xc+7e/+qgyDxrgwrdavApcXl6e87mgpqbsdDkAbaIiPG9rLCcoKJBW4eGNXZZP9LshBZvNxq5vXO98N5vN7Nv3A/36KS8I+wOHxaJY/JVXgVuwYAEHDx50fq6srKRv37588803jVaYGmw2G6/9ezWBLVrwlz/fVue2/92+i+8PHiZt0C3o9co7xpuCuGtiKS+vcHuD7smSUqKj2xAUFOSDyjyw2pSLn7qqIaXD4aC6uhqrhk7HuvPC4tfZd6CQzIlj6dKp/WW3O3biJE88l0NsdBsez5qoYoXqCg0JwWx2f9dGTc35G4RDQ72bLEdVNpty8VPN5rLApV5e+h/e+mA9o9KHMGlcxmW3Ky4p5YH/eQKdTserC54jKtLz0FOrqs+dI8bNcS5AcLD+/Db+eGnAj4eQl2o2J00utmTZSl7PX82Iv/yZZx6//NO/J0+VcX/236muPkfeon/R/douKlapvlMlZbRtG+X2Icv4a9phNP6KxQ9/uR1Wm2LxV173cKdOnXLec1ZVVQVAcXGxs+1SF+aC8DdLlq3k1eWrSB+SyuyZU9DpdG63O3mqjPuyZnDmbDV5i54nqft1Kleqvm9372Pw4Fu56cbefPnVLme7Xq8nJSWZL7742ofV1cGPh5CX8jpwCxcuZOHChS5tzzzzjGI7h8OBTqfzelIVNb26fBWvLl/FsLRBPPfk1MteSyspPd+zVZ05y9JF/yI5sZvKlfrGu++tY+bfs3n00YkugZv4wD2EhYXy1ttrfVhdHWq1cy7Bq8DNmeOfV+2vxOoP1rNk2UriYmP4Y7/ebPi/z13Wt4mMYMBNfTl7tpr7s2dy8lQZ99w1nKPHT3L0uOuF1/439qFtVKSK1avjwIFDvPLqv8nKvJ/33s1j48YtJCV2IyvrfrZt287q1X4auKbWw7mbjEVrDhT+CMCpstM89c8FivX9+lzPgJv6UmmqorikFIC33l/n9ruWv/xCkwwcwLTpz3LsWDETJ97L0CGDKC+vYMmSFTw7a75Pbuvyhj8fs12q2dzapSXyeI4rT7d2nZk2XNEW/qL7P5a+1mwvC4imw2G1+7oEr0nghPY1tZMmQvgzh016OCFUI0NKIVTkqJXACaEeq9+caPdIAic0zyGBE0I99loJnBCqcWjnqoAETmifBE4IFdkt7h+x8kcSOKF5dqsETgjV2G0SOCFUY5MhpRDqsVu1MzWPBE5onk0CJ4R6bHIMJ4R67Dbp4YRQjVWGlEKox26XIeVVie1yh69L8Av5bet+sYhwZbNrp4fTTqVCXIbVFqBYrtTp06fJyclh7Nix9OnTh4SEBHbu3Ol2282bNzNy5Eiuv/56br31VnJzc71+sY0ETmiezaFTLFfql19+IS8vj7KyMhISEi673bZt28jMzKR169Y8/fTTpKamsmTJEq8nS/arIaUQV6MhhpTJycl8/fXXREZGsmnTJjIzM91uN2/ePHr06MGyZcto0aIFAGFhYSxdupSxY8fSuXPnOvcjPZzQPItDp1iuVHh4OJGRdc+mfeTIEY4cOUJGRoYzbAD33HMPdrudzz77zON+pIcTmmdz02+YTCZMJpOi3WAwYDAYrmo/F94C3LNnT5f22NhY2rVr5/KW4MuRwAnNs6Hs0fLz88nNzVW0Z2VlkZ19+XcC1sVoNAIQHR2tWBcdHc3p06c9focETmiexU3gxo8f7/YlNFfbuwHU1NQAuH1hpV6v59w5z2+HlcAJzbO6ealmfYaOlxMcHAxAba3yPehms9m5vi5y0kRons3N0hguDCUvDC0vZjQaiYmJ8fgdEjiheRadTrE0hqSkJAAOHDjg0l5WVkZpaalzfV0kcELzrDqdYmkM3bp1o2vXrrzzzjvYLnrr6urVqwkICGDw4MEev0OO4YTmNdQMC6+88goARUVFABQUFLB7924MBgNjxowBYMaMGTzyyCM88MADDB06lB9//JFVq1aRkZFBly5dPO7Dr96AGtWqeby83pOXW/3B1yX4lXtLVta5fln7MYq2B4rr/hl3LndLV3x8PFu2bHF+3rRpE7m5uRQVFREVFcWdd97J5MmTCQz03H9JDyc0r6HmgT18+LBX26WmppKamnpV+5DACc3T0KRdEjihfRqa0kQCJ7RPQ68WkMAJ7ZMhpRAqsuE3J9o9ksAJzWusW7kaQ7MP3JTpD5GSkkxK75507tKB48eK6d2z+U7iE9zWQK/H7uSa1N4Et21NjbGSExt3sz/nAyymal+X51atron0cCUlJURFRXl1F7RWPfO/j1FR8Rv79x6kdetWvi7Hp/RtDNyxYRYhsZEcWbmFykPFRCS2p9u4QcT8MYHP0mdjO6e8U97XmkwPN2jQIObNm8ewYcPUqkd1fa6/nWNHTwDw1c4NhIWF+rgi3+n56HDCO0Tz5eQlHPtwh7Pd+O1PDHwlk6QHh3BgcYEPK3RPS8dwdd687Ed3fTWaC2ETEDsgCes5s0vYAI4VfI31XC1dM272UWV1s+BQLP5KnhYQTgH6IGw1FuUKhwNbTS2tOseijwpXvzAPbDgUi7/yGDhdIz3qIPzP74eL0UeGE5nc0aU9Mrkj+sjzQQuNb+uL0uqkpR7O41nK559/noULF3r1ZTqdjk2bNtW7KOEbh/I+pX1aPwa+ls3uZ1dSebiYiO7x3DB7LLZaKy1aBhIYopzPw9esfhywS3kMXFxcHO3atVOjFuFjxl2H+eqRXPo9N47bVj4OgN1qo+itz9G3NdBx6I1YqjxPlKM2fx5CXspj4CZMmNCkz1IKV8c/2sWJj78hIqkDgWEhmIpOYf7VxB0bZmG3WKk6WubrEhUsDruvS/Bas7/wLZQcdge//XDc+Tk4ujVRPTtR9vUhP70O14R6ONHM6XT0e24cuhYB/OCH1+BAAqcpo+9Op0OHeADatI2iZVAQ0x+fDMCJEyd5923//CVrDIGhetI+ns2Jjd9y5oSRoFYhdB7RnzYpXdk7513Kthf6ukS3LDSRIeWhQ4fUqsNnxowbxcA/uc4h8tQzUwH48oudzSpwdouV3w4ep/PI/oTERGA9V8uv+35my99e4NS2731d3mXZ5BhOO4YPVU5A01zZLTa+mrzE12VcMRlSCqEiOUsphIpsTeUYTggtkGM4IVRklcAJoR6rDCmFUI8MKYVQkdWhnUkWJHBC86SHE0JF0sMJoSLp4YRQkQROCBVZ7dp5nYcETmie9HBCqEgCJ4SKrHbtnKWUiWCF5tkcdsVypWpra5k/fz4DBw6kV69ejB49mh07dnj+wSskgROaZ7PbFcuVmjlzJvn5+QwfPpynnnqKgIAAJk2axHfffdegtUrghOZZ7DbFciX279/Phg0beOyxx5gxYwYZGRnk5+cTFxdHTk5Og9YqgROaV98h5SeffEJQUBCjRo1ytun1eu666y52797N6dOnG6xWOWkiNM/uJmAmkwmTyaRoNxgMGAwGl7bCwkK6dOlCWFiYS3uvXr1wOBwUFhYSExPTILX6VeAqqn7ydQlCg8w1yleOvfzyy+Tm5iras7KyyM7OdmkzGo3ExsYqto2OjgaQHk4IT8aPH8/IkSMV7Zf2bgA1NTUEBQUp2vV6PQBms7nB6pLAiSbJ3dDxcoKDg7FYlO/FuxC0C8FrCHLSRDR70dHRboeNRqMRoMGO30ACJwSJiYn88ssvnD171qV93759zvUNRQInmr20tDQsFgvvvfees622tpY1a9bQt29ftydUrpYcw4lmLyUlhbS0NHJycjAajXTs2JG1a9dSUlLCnDlzGnRfOofDoZ2J2YVoJGazmUWLFrF+/Xp+//13EhISmDZtGgMGDGjQ/UjghFCRHMMJoSIJnBAqkpMmwJo1a3jiiSfcrps+fToPPvigyhWpLyEhwavtNm/eTPv27Ru5mqZLAneRqVOnEhcX59LWo0cPH1Wjrnnz5rl8zs/Pp6SkRPGHKCoqSs2ymhwJ3EVuueUWkpKSfF2GT6Snp7t8/vTTT6msrFS0i/qRYzghVCQ93EVMJhMVFRXOzzqdjsjISB9WJJoaCdxFxo0b5/I5NDS0wee0EM2bBO4is2bNomPHjs7PLVq08GE1oimSwF0kJSWl2Z40EeqQkyZCqEgCJ4SKJHBCqEgCJ4SKJHBCqEiehxNCRdLDCaEiCZwQKpLACaEiCZwQKpLACaEiCZwQKpLACaEiCZwQKpLACaEiCZwQKvp/praND0MrZr0AAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAACYCAYAAACPk4f7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAARUUlEQVR4nO3deVQUV74H8G+j7NoC2iwuiLggLiDoZOISYyLxoI4iidqOijouiQngKGZ8ZoyJOpnoU4xxxCTKqIMJQaPCIHqMBnlxMu5LxBAJPhBcgkDzDGkUe6/3B2NLUy3dCNyqgt/nnDonfavs+uUcv95bt6tuyTiO40AIYcJB6AIIaUsocIQwRIEjhCEKHCEMUeAIYYgCRwhDFDhCGGovdAGECO3atWvIyMjA+fPnUVpaCg8PD4SFhWHp0qXo2bOnxbFXrlzBpk2bcP36dXTo0AHjx4/H8uXL4erqate5ZGL64VtfeVPoEkTBtesLQpcgKgbdzw3ut/b3xrFLoN3fv2TJEly5cgWRkZEICgqCSqVCamoqampqcPDgQfTu3RsAkJ+fD6VSiT59+mDatGkoKyvD7t27MXLkSHz22Wd2nYt6OCJ9em2T/vi8efOQmJgIJycnc9uECRMwadIkJCcnY8OGDQCAjz76CB4eHvj888/h7u4OAOjevTveffddnD17FsOHD7d5LrqGI5LHGQ28rTHCw8MtwgYAAQEB6Nu3L4qKigAADx48wJkzZzBlyhRz2AAgKioKbm5uOHbsmF3nosAR6TNo+VsTcRyHyspKeHp6AgAKCgpgMBgwaNAgi+OcnJwQHByM/Px8u76XhpRE8qz1aGq1Gmq1mtcul8shl8ttfufhw4dRXl6OZcuWAQBUKhUAQKFQ8I5VKBS4evWqXbVS4Ij0WQlcSkoKkpKSeO1xcXGIj49v8OuKioqwbt06DB06FFFRUQAAjUYDALyhJwA4Ozub99tCgSPSZ2XSZO7cuYiOjua12+rdVCoV3njjDXTq1Albt26Fg0PtVZeLiwsAQKfT8f6MVqs177eFAkekz0oPZ+/Qsa7q6mosWrQI1dXVSEtLsxg+Pv7vx0PLulQqFby9ve06B02aEMnjTHre1lharRaLFy9GSUkJduzYgcBAy9/x+vXrh/bt2yMvL8+iXafTIT8/H8HBwXadhwJHpE+v5W+NYDQasXTpUly9ehVbt27FkCFDeMd07NgRw4cPR2ZmJh4+fGhuz8zMRE1NDSIjI+06Fw0pifQ18ne3+jZs2ICcnBy89NJLqKqqQmZmpnmfu7s7IiIiAADLli3DjBkzEBMTY77TZM+ePRg9ejRGjBhh17no1i4Rolu7LNm6tUtzOpXX5jJylt3fHxMTgwsXLljd161bN+Tk5Jg/X7p0CYmJieZ7KSdMmICEhAS4ubnZdS4KnAhR4CzZCtyjb3fz2lzHzG+pcpqEhpRE+po4pGSJAkekjwJHCEN6/o/RYkWBI9JHPRwhDBmohyOEHQP1cKJTcvsujhzPwZkLV3Cn9B60Wj16dPPDuJdHIWZ6NNxcn9x8+sP1AmQdz8H1gkIUFN7Eo0cafPDnBEyZ+IqA/wdsyGQyLIlfiEWLZiOgZ3eoVPdx8GAW3l+7CTU1j4QuzzqjUegK7NZmbu3KOHoCe/f/Ez26+WHxvJlYHrsAAf7dsW3nXsxenACN9sntQP86exH70o+guvoBgvrYvzZGa7A5cQ02J65Bfv4N/HHpahw6dARxcfORmZECmUwmdHnW6XX8TaTaTA/3yphRWBijRMcOTx6PV0ZPRM8eXbEzZR/Ss45j5tTJ5vY/zJwKN1cXnPif73D1h+tClc3UgAH9EBc7H+kZRzFd+bq5vbjkNrZ+/AGUyijs2/dPASt8CurhxGdQcD+LsD0WOXY0AOB/b94yt3Xx8rQYYrYVM5RT4ODggL/97e8W7X/f9SUePqzBrN+/KlBlNhgN/E2k7ApccnKyeTGV1qa8ohIA0NnLQ+BKhDdsaCiMRiMuXLRcLkCr1SI390cMG8a/i14MOL2et4mVXYHbvHkzrl9/MqyqqqpCeHg4Ll682GKFsWA0GvHZP9LQvl07THzlJaHLEZxfVx9UVt63+lTzz6VlUCg6w9HRUYDKbDAY+ZtIPdOQkuM41NTUwCCh6Vhr/nvrDuTm5SN2YQx69ewudDmCc3N1hVZrfcJBo6mdVHJzs2+FYaaMRv4mUm1m0qS+bTv34stDWZgWNR6L5iiFLkcUah49greV61wAcHFxrj1GjD8NiHgIWV+bmTSpa/uuL7AjJQ1TJr6C9/7U8ApObcm90nJ06eJldWWqbl19oVL9H/Qi/MvNGYy8Tazs7uHu3buHn376CUDtYisAcPfuXXNbff3792+G8prf9l1f4NPdqYgaH4F1K5eK97clAVy6nItx48bgud8Mwb9PP3kg09nZGaGhA/Hdd+cErK4BIh5C1md34LZs2YItW7ZYtL333nu84ziOg0wms3slWpY+3Z2KT3enYlLkWPzlz8vMS6CRWl8dOIyV/xWPJUsWWgRu4YKZcHd3w5f7MgSsrgE66cwl2BW49evXt3QdLS7tUBa27/oCfj7eeH7YEBz95luL/Z09PTDiuXAAQGlZObK+rn2svrC49ve5b0+fR7mq9ieESZEvo6uvD7viGcnL+wmffPoPxMXOx4GvknHsWA6C+/dFXNx8nDp1BmlpIg1ca+vhrC2oKTV5+TcAAPfKK7Dqg828/cPCBpsDd7e0HNuS91rszz51GtmnTgMAwkIGtsrAAUDC8vdx69ZdLFw4CxPGj0Vl5X1s374H76/dBBGtxmFBzNds9dGaJiJEa5pYsrWmyYOEyby2Dh8dbqlymqTN/ixAWg/OYBK6BLtR4Ij0tbZJE0LEjDNSD0cIMzSkJIQhTkeBI4Qdg2gm2m2iwBHJ4yhwhLBj0lHgCGGGk86vAhQ4In0UOEIYMuml84gVBY5InslAgSOEGZORAkcIM0YaUhLCjskgnSf3KXBE8owUOELYMdI1HCHsmIzS6eGkUykhT2EwOPC2xqqoqEBiYiJiYmIQFhaGoKAgnD9/3uqxJ0+eRHR0NAYPHowxY8YgKSnJ7lXIKXBE8kwmGW9rrOLiYiQnJ6O8vBxBQUFPPe7UqVOIjY1Fp06dsHr1akRERGD79u12r2wnqiGlX2Ck0CWIQo7XCKFLkBSjqen9xsCBA3Hu3Dl4enoiOzsbsbGxVo/buHEjBgwYgF27dqFdu3YAAHd3d+zcuRMxMTEICAho8DzUwxHJMxgdeFtjdejQAZ6eng0eU1hYiMLCQiiVSnPYAGDmzJkwmUw4ceKEzfOIqocj5FkYOTazlI9f2TZo0CCLdh8fH/j6+lq80u1pKHBE8qwNKdVqNdRqNa9dLpdDLpc/03lUKhUAQKFQ8PYpFApUVFTY/A4KHJE8vZUeLiUlBUlJSbz2uLg4xMc/2xuTNBoNAFh9u5CzszMePbL9Ki8KHJE8o5WpiLlz51pdov9ZezcAcHGpfe+7tTfEarVa8/6GUOCI5BnB7+GaMnR8msdDSZVKBW9vb4t9KpUKYWFhNr+DZimJ5Okh420tITg4GACQl5dn0V5eXo6ysjLz/oZQ4IjkGWQy3tYS+vbti8DAQOzfvx/GOq/ISktLg4ODA8aNG2fzO2hISSSvuV5W9cknnwAAioqKAACZmZm4fPky5HI5Zs+eDQBYsWIF3nzzTSxYsAATJkzAjRs3kJqaCqVSiV69etk8h6heV9VF3k/oEkQh3S1U6BJEZXTZgQb37/ebxWtT3ktt9HmedktXt27dkJOTY/6cnZ2NpKQkFBUVwcvLC6+99hreeusttG9vu/+iHo5IXnMNIQsKCuw6LiIiAhEREc90DgockTwJrbBAgSPSJ6HnTylwRPoktA4sBY5IHw0pCWGIhpSEMERDSkIYoiElIQwZIZp7N2yiwBHJa65bu1ho84H7Y8IbCAkdgNAhgxDQqwdu37qL8MEvC11Wi+sRPwUdQgLRISQQrj19oLlTgQu/sb5wDgB0DOuDgHd+j47hfQGOg/riDRT/NRUPfyxhV/RT6GTS6eEafFqgtLTU/JRra7V6zXK88OLzKCm+jV9+qRK6HGZ6rZoFj5GDoCkph/6XBw0e2zG8L0Iz1sLF3xu3Nu7HrU1fwTXQF6GZ6+DW359RxU9ntLKJVYOBGzt2LL755htWtQhiaMhY9Av4LaZO+QPK7tlek6K1uPBcLM4OmI8flH+Brvx+g8f2/ut8mPQG5E55Hz/vPIqfdx5F7pT3AY5D7zVzGFX8dEZwvE2sGgyciB4kaDG3Su4IXYIgNLft+8fFJcAX8rA+qMw6B13Zk2Dqyu6jMuscPEYPhqPCo6XKtIseHG8TK3oAlTSo45DeAAD1Jf6d9OrLNyBzcEDH0EDWZVmQUg9nc9JE1kJPzxJpcPKtXRxVW8Yfdj5uc/L1YlpTfWLu0eqzGbgPP/wQW7ZssevLZDIZsrOzm1wUEY92rs4AAE7Lv5+D0+j/cwx/2TiWDK0pcH5+fvD19WVRCxEh4yMtAEDmzP+rInNx/M8x/GXjWBLzELI+m4GbN28eJk2axKIWIkK6sl8AAM5Who2P23RWhpss6TmToOdvDJo0IQ2qvlq7oI58GH+9D/nQfuBMJlTn3mRdlgUpTZpQ4EiDNCVlqL5aiC6TnoeTz5O3yzj5eKLLpOdR9e886FXC3jAgpcC1+Vu7ps2IQo8eXQEAnbt4wcnREQl/ehMAcOdOKQ7syxSyvBbjPXU0XLp3AQA4dpZD5tge/ktfBQBo7lai4uC/zMcWvbsHIYfWIDRzHUp3HQMAdF0wHjIHB9xcs5d98fXoIZ0hZZtfJi/z6OcY+cJvre47/d15RE2MYVwRm2XyQtLXwGPEQKv7qs78iGuvrrFo6zi0HwJWzoA8vC84joP6YgFKPvwSD34obvFabS2TF+X/O15b5u0jLVVOk7T5wIkRrUtpyVbgfuc/kdd25PbRliqnSdr8kJJIn5RmKSlwRPKMErqGo8ARyTNSD0cIOwYKHCHsGGhISQg7NKQkhCEDJ+ZFFSxR4IjkUQ9HCEPUwxHCEPVwhDBEgSOEIYNJOq/zoMARyaMejhCGKHCEMGQwSWeWkpZYIJJn5Ey8rbF0Oh02bdqEUaNGISQkBNOnT8fZs2ebvVYKHJE8o8nE2xpr5cqVSElJweTJk7Fq1So4ODhg0aJF+P7775u1VgockTy9ycjbGuPatWs4evQo3n77baxYsQJKpRIpKSnw8/NDYmJis9ZKgSOS19Qh5ddffw1HR0dMmzbN3Obs7IypU6fi8uXLqKhovrcq0aQJkTyTlYCp1Wqo1Wpeu1wuh1wut2jLz89Hr1694O7ubtEeEhICjuOQn58Pb2/vZqlVVIGrVN8QugQiQVoN/5Vj27ZtQ1JSEq89Li4O8fHxFm0qlQo+Pj68YxUKBQBQD0eILXPnzkV0dDSvvX7vBgAajQaOjo68dmfn2heZaLXaZquLAkdaJWtDx6dxcXGBXq/ntT8O2uPgNQeaNCFtnkKhsDpsVKlUANBs128ABY4Q9O/fH8XFxXj48KFFe25urnl/c6HAkTYvMjISer0eBw48WeFZp9MhPT0d4eHhVidUnhVdw5E2LzQ0FJGRkUhMTIRKpYK/vz8yMjJQWlqK9evXN+u5RPVuAUKEotVq8fHHHyMrKwu//vorgoKCkJCQgBEjRjTreShwhDBE13CEMESBI4QhmjQBkJ6ejnfeecfqvuXLl+P1119nXBF7QUH8d3hbc/LkSXTv3r2Fq2m9KHB1LFu2DH5+fhZtAwYMEKgatjZu3GjxOSUlBaWlpbx/iLy8vFiW1epQ4Op48cUXERwcLHQZgoiKirL4fPz4cVRVVfHaSdPQNRwhDFEPV4darcb9+/fNn2UyGTw9PQWsiLQ2FLg65syZY/HZzc2t2de0IG0bBa6OtWvXwt/f3/y5Xbt2AlZDWiMKXB2hoaFtdtKEsEGTJoQwRIEjhCEKHCEMUeAIYYgCRwhD9DwcIQxRD0cIQxQ4QhiiwBHCEAWOEIYocIQwRIEjhCEKHCEMUeAIYYgCRwhDFDhCGPp/ZA5xYg2FwLsAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 216x144 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ }
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "KrI1Zz1pjKQ7",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "b207b790-37d4-4848-d555-6ad220adaa7b"
+ },
+ "source": [
+ "# Thus in binary classification, the count of true negatives is C0.0, false negatives is C1.0, true positives is C1.1 and false positives is C0.1.\n",
+ "len(confusion_matrix_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "76"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 106
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "BsC-jVfPjKQ9"
+ },
+ "source": [
+ "# importing mean() for calculate the mean of all the lists (precision, recall, f1-score ... ) \n",
+ "from statistics import mean \n",
+ " "
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "rdzvtIa0jKQ-",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "d555f39d-c253-4a1d-b603-9956d442d833"
+ },
+ "source": [
+ "mean(roc_auc_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9450712322479564"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 108
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "w9bvZAo8jKQ_",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "63ac6f3b-7fbe-41bc-c9f3-96678f46fe30"
+ },
+ "source": [
+ "mean(recall_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.8910835144387776"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 109
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "qEdm--cNjKRA",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "1ae955f7-8399-431b-9d96-021cc56559b9"
+ },
+ "source": [
+ "mean(accuracy_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9703947368421053"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 110
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "-iwsZ8oLjKRL",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "a1076a0b-df4e-4689-8eac-dbd4e7bfd737"
+ },
+ "source": [
+ "mean(precision_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9929824561403509"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 111
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "YFi-fYAtjKRR",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "dd971dd5-b18b-4761-c178-7cd4319c5a98"
+ },
+ "source": [
+ "mean(f1_score_met)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9354850323637031"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 112
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "nHZI8T7wjKRT",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "9649b7ec-6e98-4494-e1bd-43415d57d6f8"
+ },
+ "source": [
+ "mean(matthews_set)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "0.9212690080028949"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 113
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "jbQz1oszjKRU",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 36
+ },
+ "outputId": "44a13933-7085-4bd4-e317-233a4bf7947b"
+ },
+ "source": [
+ "# Combine the predictions for each batch into a single list of 0s and 1s.\n",
+ "flat_predictions = [item for sublist in predictions_test for item in sublist]\n",
+ "flat_predictions = np.argmax(flat_predictions, axis=1).flatten()\n",
+ "\n",
+ "# Combine the correct labels for each batch into a single list.\n",
+ "flat_true_labels = [item for sublist in true_labels for item in sublist]\n",
+ "\n",
+ "# Calculate the MCC\n",
+ "mcc = matthews_corrcoef(flat_true_labels, flat_predictions)\n",
+ "\n",
+ "print('MCC: %.3f' % mcc)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "MCC: 0.921\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "rbvvknVbjKRW"
+ },
+ "source": [
+ "#5. Test sample of novels sentences"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "SRm_cTWfjKRX"
+ },
+ "source": [
+ "##Download the novels"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "zXhvrtXajKRY"
+ },
+ "source": [
+ "We must create a new folder in (content) on colab and we name it (Romans). Then we upload the novels we want"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "127M_YhdjKRY"
+ },
+ "source": [
+ "from os import listdir\n",
+ "from os.path import isfile, join\n",
+ "dir = \"/content/Romans\"\n",
+ "onlyfiles = [f for f in listdir(dir) if isfile(join(dir, f))] # list of novels names(with extentions filename.txt) in the directory path"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "mwecaybtjKRd",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "5af26370-bf86-4fa6-f2b1-e15d0cc5ce9b"
+ },
+ "source": [
+ "onlyfiles[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "'maison.txt'"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 51
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "6PnRsMxPjKRk"
+ },
+ "source": [
+ "# Put the novels contents in list\n",
+ "content_french = [] #List of lists of sentences of novels\n",
+ "file_content =[]\n",
+ "for file in onlyfiles:\n",
+ " #file_content.append( file.read )\n",
+ " f = open('/content/Romans/'+ file)\n",
+ " #file_content.append(f)\n",
+ " content_french.append(f.read())"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "lPnJCdMTjKRl",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 55
+ },
+ "outputId": "24ea0245-f063-4682-e479-d2539b6d95b7"
+ },
+ "source": [
+ "content_french[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "'\\n\\n\\nAu milieu de la rue Saint-Denis, presque au coin de la rue du Petit-Lion, existait naguère une de ces maisons précieuses qui donnent aux historiens la facilité de reconstruire par analogie l’ancien Paris. Les murs menaçants de cette bicoque semblaient avoir été bariolés d’hiéroglyphes. Quel autre nom le flâneur pouvait-il donner aux X et aux V que traçaient sur la façade les pièces de bois transversales ou diagonales dessinées dans le badigeon par de petites lézardes parallèles ? Evidemment, au passage de toutes les voitures, chacune de ces solives s’agitait dans sa mortaise. Ce vénérable édifice était surmonté d’un toit triangulaire dont aucun modèle ne se verra bientôt plus à Paris. Cette couverture, tordue par les intempéries du climat parisien, s’avançait de trois pieds sur la rue, autant pour garantir des eaux pluviales le seuil de la porte, que pour abriter le mur d’un grenier et sa lucarne sans appui. Ce dernier étage était construit en planches clouées l’une sur l’autre comme des ardoises, afin sans doute de ne pas charger cette frêle maison.\\n\\nPar une matinée pluvieuse, au mois de mars, un jeune homme, soigneusement enveloppé dans son manteau, se tenait sous l’auvent de la boutique qui se trouvait en face de ce vieux logis, et paraissait l’examiner avec un enthousiasme d’archéologue. A la vérité, ce débris de la bourgeoisie du seizième siècle pouvait offrir à l’observateur plus d’un problème à résoudre. Chaque étage avait sa singularité. Au premier, quatre fenêtres longues, étroites, rapprochées l’une de l’autre, avaient des carreaux de bois dans leur partie inférieure, afin de produire ce jour douteux, à la faveur duquel un habile marchand prête aux étoffes la couleur souhaitée par ses chalands. Le jeune homme semblait plein de dédain pour cette partie essentielle de la maison, ses yeux ne s’y étaient pas encore arrêtés. Les fenêtres du second étage, dont les jalousies relevées laissaient voir, au travers de grands carreaux en verre de Bohême, de petits rideaux de mousseline rousse, ne l’intéressaient pas davantage. Son attention se portait particulièrement au troisième, sur d’humbles croisées dont le bois travaillé grossièrement aurait mérité d’être placé au Conservatoire des arts et métiers pour y indiquer les premiers efforts de la menuiserie française. Ces croisées avaient de petites vitres d’une couleur si verte, que, sans son excellente vue, le jeune homme n’aurait pu apercevoir les rideaux de toile à carreaux bleus qui cachaient les mystères de cet appartement aux yeux des profanes. Parfois, cet observateur, ennuyé de sa contemplation sans résultat, ou du silence dans lequel la maison était ensevelie, ainsi que tout le quartier, abaissait ses regards vers les régions inférieures. Un sourire involontaire se dessinait alors sur ses lèvres, quand il revoyait la boutique où se rencontraient en effet des choses assez risibles. Une formidable pièce de bois, horizontalement appuyée sur quatre piliers qui paraissaient courbés par le poids de cette maison décrépite, avait été rechampie d’autant de couches de diverses peintures que la joue d’une vieille duchesse en a reçu de rouge. Au milieu de cette large poutre mignardement sculptée se trouvait un antique tableau représentant un chat qui pelotait. Cette toile causait la gaieté du jeune homme. Mais il faut dire que le plus spirituel des peintres modernes n’inventerait pas de charge si comique. L’animal tenait dans une de ses pattes de devant une raquette aussi grande que lui, et se dressait sur ses pattes de derrière pour mirer une énorme balle que lui renvoyait un gentilhomme en habit brodé. Dessin, couleurs, accessoires, tout était traité de manière à faire croire que l’artiste avait voulu se moquer du marchand et des passants. En altérant cette peinture naïve, le temps l’avait rendue encore plus grotesque par quelques incertitudes qui devaient inquiéter de consciencieux flâneurs. Ainsi la queue mouchetée du chat était découpée de telle sorte qu’on pouvait la prendre pour un spectateur, tant la queue des chats de nos ancêtres était grosse, haute et fournie. A droite du tableau, sur un champ d’azur qui déguisait imparfaitement la pourriture du bois, les passants lisaient Guillaume ; et à gauche, Successeur du sieur Chevrel. Le soleil et la pluie avaient rongé la plus grande partie de l’or moulu parcimonieusement appliqué sur les lettres de cette inscription, dans laquelle les U remplaçaient les V et réciproquement, selon les lois de notre ancienne orthographe. Afin de rabattre l’orgueil de ceux qui croient que le monde devient de jour en jour plus spirituel, et que le moderne charlatanisme surpasse tout, il convient de faire observer ici que ces enseignes, dont l’étymologie semble bizarre à plus d’un négociant parisien, sont les tableaux morts de vivants tableaux à l’aide desquels nos espiègles ancêtres avaient réussi à amener les chalands dans leurs maisons. Ainsi la Truie-qui-file, le Singe-vert, etc., furent des animaux en cage dont l’adresse émerveillait les passants, et dont l’éducation prouvait la patience de l’industriel au quinzième siècle. De semblables curiosités enrichissaient plus vite leurs heureux possesseurs que les Providence, les Bonne-foi, les Grâce-de-Dieu et les Décollation de saint Jean-Baptiste qui se voient encore rue Saint-Denis. Cependant l’inconnu ne restait certes pas là pour admirer ce chat, qu’un moment d’attention suffisait à graver dans la mémoire. Ce jeune homme avait aussi ses singularités. Son manteau, plissé dans le goût des draperies antiques, laissait voir une élégante chaussure, d’autant plus remarquable au milieu de la boue parisienne, qu’il portait des bas de soie blancs dont les mouchetures attestaient son impatience. Il sortait sans doute d’une noce ou d’un bal car à cette heure matinale il tenait à la main des gants blancs et les boucles de ses cheveux noirs défrisés éparpillées sur ses épaules indiquaient une coiffure à la Caracalla, mise à la mode autant par l’Ecole de David que par cet engouement pour les formes grecques et romaines qui marqua les premières années de ce siècle. Malgré le bruit que faisaient quelques maraîchers attardés passant au galop pour se rendre à la grande halle, cette rue si agitée avait alors un calme dont la magie n’est connue que de ceux qui ont erré dans Paris désert, à ces heures où son tapage, un moment apaisé, renaît et s’entend dans le lointain comme la grande voix de la mer. Cet étrange jeune homme devait être aussi curieux pour les commerçants du Chat-qui-pelote, que le Chat-qui-pelote l’était pour lui. Une cravate éblouissante de blancheur rendait sa figure tourmentée encore plus pâle qu’elle ne l’était réellement. Le feu tour à tour sombre et pétillant que jetaient ses yeux noirs s’harmoniait avec les contours bizarres de son visage, avec sa bouche large et sinueuse qui se contractait en souriant. Son front, ridé par une contrariété violente, avait quelque chose de fatal. Le front n’est-il pas ce qui se trouve de plus prophétique en l’homme ? Quand celui de l’inconnu exprimait la passion, les plis qui s’y formaient causaient une sorte d’effroi par la vigueur avec laquelle ils se prononçaient ; mais lorsqu’il reprenait son calme, si facile à troubler, il y respirait une grâce lumineuse qui rendait attrayante cette physionomie où la joie, la douleur, l’amour, la colère, le dédain éclataient d’une manière si communicative que l’homme le plus froid en devait être impressionné. Cet inconnu se dépitait si bien au moment où l’on ouvrit précipitamment la lucarne du grenier, qu’il n’y vit pas apparaître trois joyeuses figures rondelettes, blanches, roses, mais aussi communes que le sont les figures du Commerce sculptées sur certains monuments. Ces trois faces, encadrées par la lucarne, rappelaient les têtes d’anges bouffis semés dans les nuages qui accompagnent le Père éternel. Les apprentis respirèrent les émanations de la rue avec une avidité qui démontrait combien l’atmosphère de leur grenier était chaude et méphitique. Après avoir indiqué ce singulier factionnaire, le commis qui paraissait être le plus jovial disparut et revint en tenant à la main un instrument dont le métal inflexible a été récemment remplacé par un cuir souple ; puis tous prirent une expression malicieuse en regardant le badaud qu’ils aspergèrent d’une pluie fine et blanchâtre dont le parfum prouvait que les trois mentons venaient d’être rasés. Elevés sur la pointe de leurs pieds, et réfugiés au fond de leur grenier pour jouir de la colère de leur victime, les commis cessèrent de rire en voyant l’insouciant dédain avec lequel le jeune homme secoua son manteau, et le profond mépris que peignit sa figure quand il leva les yeux sur la lucarne vide. En ce moment, une main blanche et délicate fit remonter vers l’imposte la partie inférieure d’une des grossières croisées du troisième étage, au moyen de ces coulisses dont le tourniquet laisse souvent tomber à l’improviste le lourd vitrage qu’il doit retenir. Le passant fut alors récompensé de sa longue attente. La figure d’une jeune fille, fraîche comme un de ces blancs calices qui fleurissent au sein des eaux, se montra couronnée d’une ruche en mousseline froissée qui donnait à sa tête un air d’innocence admirable. Quoique couverts d’une étoffe brune, son cou, ses épaules s’apercevaient, grâce à de légers interstices ménagés par les mouvements du sommeil. Aucune expression de contrainte n’altérait ni l’ingénuité de ce visage, ni le calme de ces yeux immortalisés par avance dans les sublimes compositions de Raphaël : c’était la même grâce, la même tranquillité de ces vierges devenues proverbiales. Il existait un charmant contraste produit par la jeunesse des joues de cette figure, sur laquelle le sommeil avait comme mis en relief une surabondance de vie, et par la vieillesse de cette fenêtre massive aux contours grossiers, dont l’appui était noir. Semblable à ces fleurs de jour qui n’ont pas encore au matin déplié leur tunique roulée par le froid des nuits, la jeune fille, à peine éveillée, laissa errer ses yeux bleus sur les toits voisins et regarda le ciel ; puis, par une sorte d’habitude, elle les baissa sur les sombres régions de la rue, où ils rencontrèrent aussitôt ceux de son adorateur. La coquetterie la fit sans doute souffrir d’être vue en déshabillé, elle se retira vivement en arrière, le tourniquet tout usé tourna, la croisée redescendit avec cette rapidité qui, de nos jours, a valu un nom odieux à cette naïve invention de nos ancêtres, et la vision disparut. Il semblait à ce jeune homme que la plus brillante des étoiles du matin avait été soudain cachée par un nuage.\\n\\nPendant ces petits événements, les lourds volets intérieurs qui défendaient le léger vitrage de la boutique du Chat-qui-pelote avaient été enlevés comme par magie. La vieille porte à heurtoir fut repliée sur le mur intérieur de la maison par un serviteur vraisemblablement contemporain de l’enseigne, qui d’une main tremblante y attacha le morceau de drap carré sur lequel était brodé en soie jaune le nom de Guillaume, successeur de Chevrel. Il eût été difficile à plus d’un passant de deviner le genre de commerce de monsieur Guillaume. A travers les gros barreaux de fer qui protégeaient extérieurement sa boutique, à peine y apercevait-on des paquets enveloppés de toile brune aussi nombreux que des harengs quand ils traversent l’Océan. Malgré l’apparente simplicité de cette gothique façade, monsieur Guillaume était, de tous les marchands drapiers de Paris, celui dont les magasins se trouvaient toujours le mieux fournis, dont les relations avaient le plus d’étendue, et dont la probité commerciale était la plus exacte. Si quelques-uns de ses confrères avaient conclu des marchés avec le gouvernement, sans avoir la quantité de drap voulue, il était toujours prêt à la leur livrer, quelque considérable que fût le nombre de pièces soumissionnées. Le rusé négociant connaissait mille manières de s’attribuer le plus fort bénéfice sans se trouver obligé, comme eux, de courir chez des protecteurs, y faire des bassesses ou de riches présents. Si les confrères ne pouvaient le payer qu’en excellentes traites un peu longues, il indiquait son notaire comme un homme accommodant ; et savait encore tirer une seconde mouture du sac, grâce à cet expédient qui faisait dire proverbialement aux négociants de la rue Saint-Denis : — Dieu vous garde du notaire de monsieur Guillaume ! pour désigner un escompte onéreux. Le vieux négociant se trouva debout comme par miracle, sur le seuil de sa boutique, au moment où le domestique se retira. Monsieur Guillaume regarda la rue Saint-Denis, les boutiques voisines et le temps, comme un homme qui débarque au Havre et revoit la France après un long voyage. Bien convaincu que rien n’avait changé pendant son sommeil, il aperçut alors le passant en faction, qui, de son côté, contemplait le patriarche de la draperie, comme Humboldt dut examiner le premier gymnote électrique qu’il vit en Amérique. Monsieur Guillaume portait de larges culottes de velours noir, des bas chinés, et des souliers carrés à boucles d’argent. Son habit à pans carrés, à basques carrées, à collet carré, enveloppait son corps légèrement voûté d’un drap verdâtre garni de grands boutons en métal blanc mais rougis par l’usage. Ses cheveux gris étaient si exactement aplatis et peignés sur son crâne jaune, qu’ils le faisaient ressembler à un champ sillonné. Ses petits yeux verts, percés comme avec une vrille, flamboyaient sous deux arcs marqués d’une faible rougeur à défaut de sourcils. Les inquiétudes avaient tracé sur son front des rides horizontales aussi nombreuses que les plis de son habit. Cette figure blême annonçait la patience, la sagesse commerciale, et l’espèce de cupidité rusée que réclament les affaires. A cette époque on voyait moins rarement qu’aujourd’hui de ces vieilles familles où se conservaient, comme de précieuses traditions, les moeurs, les costumes caractéristiques de leurs professions, et restées au milieu de la civilisation nouvelle comme ces débris antédiluviens retrouvés par Cuvier dans les carrières. Le chef de la famille Guillaume était un de ces notables gardiens des anciens usages : on le surprenait à regretter le Prévôt des Marchands, et jamais il ne parlait d’un jugement du tribunal de commerce sans le nommer la sentence des consuls. C’était sans doute en vertu de ces coutumes que, levé le premier de sa maison, il attendait de pied ferme l’arrivée de ses trois commis, pour les gourmander en cas de retard. Ces jeunes disciples de Mercure ne connaissaient rien de plus redoutable que l’activité silencieuse avec laquelle le patron scrutait leurs visages et leurs mouvements, le lundi matin, en y recherchant les preuves ou les traces de leurs escapades. Mais, en ce moment, le vieux drapier ne fit aucune attention à ses apprentis. Il était occupé à chercher le motif de la sollicitude avec laquelle le jeune homme en bas de soie et en manteau portait alternativement les yeux sur son enseigne et sur les profondeurs de son magasin. Le jour, devenu plus éclatant, permettait d’y apercevoir le bureau grillagé, entouré de rideaux en vieille soie verte, où se tenaient les livres immenses, oracles muets de la maison. Le trop curieux étranger semblait convoiter ce petit local, y prendre le plan d’une salle à manger latérale, éclairée par un vitrage pratiqué dans le plafond, et d’où la famille réunie devait facilement voir, pendant ses repas, les plus légers accidents qui pouvaient arriver sur le seuil de la boutique. Un si grand amour pour son logis paraissait suspect à un négociant qui avait subi le régime de la Terreur. Monsieur Guillaume pensait donc assez naturellement que cette figure sinistre en voulait à la caisse du Chat-qui-pelote. Après avoir discrètement joui du duel muet qui avait lieu entre son patron et l’inconnu, le plus âgé des commis hasarda de se placer sur la dalle où était monsieur Guillaume, en voyant le jeune homme contempler à la dérobée les croisées du troisième. Il fit deux pas dans la rue, leva la tête, et crut avoir aperçu mademoiselle Augustine Guillaume qui se retirait avec précipitation. Mécontent de la perspicacité de son premier commis, le drapier lui lança un regard de travers ; mais tout à coup les craintes mutuelles que la présence de ce passant excitait dans l’âme du marchand et de l’amoureux commis se calmèrent. L’inconnu héla un fiacre qui se rendait à une place voisine, et y monta rapidement en affectant une trompeuse indifférence. Ce départ mit un certain baume dans le coeur des autres commis, assez inquiets de retrouver la victime de leur plaisanterie.\\n\\n— Hé bien, messieurs, qu’avez-vous donc à rester là , les bras croisés ? dit monsieur Guillaume à ses trois néophytes. Mais autrefois, sarpejeu ! quand j’étais chez le sieur Chevrel, j’avais déjà visité plus de deux pièces de drap.\\n\\n— Il faisait donc jour de meilleure heure, dit le second commis que cette tâche concernait.\\n\\nLe vieux négociant ne put s’empêcher de sourire. Quoique deux de ces trois jeunes gens, confiés à ses soins par leurs pères, riches manufacturiers de Louviers et de Sedan, n’eussent qu’à demander cent mille francs pour les avoir, le jour où ils seraient en âge de s’établir, Guillaume croyait de son devoir de les tenir sous la férule d’un antique despotisme inconnu de nos jours dans les brillants magasins modernes dont les commis veulent être riches à trente ans : il les faisait travailler comme des nègres. A eux trois, ces commis suffisaient à une besogne qui aurait mis sur les dents dix de ces employés dont le sybaritisme enfle aujourd’hui les colonnes du budget. Aucun bruit ne troublait la paix de cette maison solennelle, où les gonds semblaient toujours huilés, et dont le moindre meuble avait cette propreté respectable qui annonce un ordre et une économie sévères. Souvent, le plus espiègle des commis s’était amusé à écrire sur le fromage de Gruyère qu’on leur abandonnait au déjeuner, et qu’ils se plaisaient à respecter, la date de sa réception primitive. Cette malice et quelques autres semblables faisaient parfois sourire la plus jeune des deux filles de Guillaume, la jolie vierge qui venait d’apparaître au passant enchanté. Quoique chacun des apprentis, et même le plus ancien, payât une forte pension, aucun d’eux n’eût été assez hardi pour rester à la table du patron au moment où le dessert y était servi. Lorsque madame Guillaume parlait d’accommoder la salade, ces pauvres jeunes gens tremblaient en songeant avec quelle parcimonie sa prudente main savait y épancher l’huile. Il ne fallait pas qu’ils s’avisassent de passer une nuit dehors, sans avoir donné long-temps à l’avance un motif plausible à cette irrégularité. Chaque dimanche, et à tour de rôle, deux commis accompagnaient la famille Guillaume à la messe de Saint-Leu et aux vêpres. Mesdemoiselles Virginie et Augustine, modestement vêtues d’indienne, prenaient chacune le bras d’un commis et marchaient en avant, sous les yeux perçants de leur mère, qui fermait ce petit cortége domestique avec son mari accoutumé par elle à porter deux gros paroissiens reliés en maroquin noir. Le second commis n’avait pas d’appointements. Quant à celui que douze ans de persévérance et de discrétion initiaient aux secrets de la maison, il recevait huit cents francs en récompense de ses labeurs. A certaines fêtes de famille, il était gratifié de quelques cadeaux auxquels la main sèche et ridée de madame Guillaume donnait seule du prix : des bourses en filet, qu’elle avait soin d’emplir de coton pour faire valoir leurs dessins à jour ; des bretelles fortement conditionnées, ou des paires de bas de soie bien lourdes. Quelquefois, mais rarement, ce premier ministre était admis à partager les plaisirs de la famille soit quand elle allait à la campagne, soit quand après des mois d’attente elle se décidait à user de son droit à demander, en louant une loge, une pièce à laquelle Paris ne pensait plus. Quant aux deux autres commis, la barrière de respect qui séparait jadis un maître drapier de ses apprentis était placée si fortement entre eux et le vieux négociant, qu’il leur eût été plus facile de voler une pièce de drap que de déranger cette auguste étiquette. Cette réserve peut paraître ridicule aujourd’hui. Néanmoins, ces vieilles maisons étaient des écoles de moeurs et de probité. Les maîtres adoptaient leurs apprentis. Le linge d’un jeune homme était soigné, réparé, quelquefois renouvelé par la maîtresse de la maison. Un commis tombait-il malade, il devenait l’objet de soins vraiment maternels. En cas de danger, le patron prodiguait son argent pour appeler les plus célèbres docteurs ; car il ne répondait pas seulement des moeurs et du savoir de ces jeunes gens à leurs parents. Si l’un d’eux, honorable par le caractère, éprouvait quelque désastre, ces vieux négociants savaient apprécier l’intelligence qu’ils avaient développée, et n’hésitaient pas à confier le bonheur de leurs filles à celui auquel ils avaient pendant long-temps confié leurs fortunes. Guillaume était un de ces hommes antiques, et s’il en avait les ridicules, il en avait toutes les qualités. Aussi Joseph Lebas, son premier commis, orphelin et sans fortune, était-il, dans son idée, le futur époux de Virginie sa fille aînée. Mais Joseph ne partageait point les pensées symétriques de son patron, qui, pour un empire, n’aurait pas marié sa seconde fille avant la première. L’infortuné commis se sentait le coeur entièrement pris pour mademoiselle Augustine la cadette. Afin de justifier cette passion, qui avait grandi secrètement, il est nécessaire de pénétrer plus avant dans les ressorts du gouvernement absolu qui régissait la maison du vieux marchand drapier.\\n\\nGuillaume avait deux filles. L’aînée, mademoiselle Virginie, était tout le portrait de sa mère. Madame Guillaume, fille du sieur Chevrel, se tenait si droite sur la banquette de son comptoir, que plus d’une fois elle avait entendu des plaisants parier qu’elle y était empalée. Sa figure maigre et longue trahissait une dévotion outrée. Sans grâces et sans manières aimables, madame Guillaume ornait habituellement sa tête presque sexagénaire d’un bonnet dont la forme était invariable et garni de barbes comme celui d’une veuve. Tout le voisinage l’appelait la soeur tourière. Sa parole était brève, et ses gestes avaient quelque chose des mouvements saccadés d’un télégraphe. Son oeil, clair comme celui d’un chat, semblait en vouloir à tout le monde de ce qu’elle était laide. Mademoiselle Virginie, élevée comme sa jeune soeur sous les lois despotiques de leur mère, avait atteint l’âge de vingt-huit ans. La jeunesse atténuait l’air disgracieux que sa ressemblance avec sa mère donnait parfois à sa figure ; mais la rigueur maternelle l’avait dotée de deux grandes qualités qui pouvaient tout contre-balancer : elle était douce et patiente. Mademoiselle Augustine, à peine âgée de dix-huit ans, ne ressemblait ni à son père ni à sa mère. Elle était de ces filles qui, par l’absence de tout lien physique avec leurs parents, font croire à ce dicton de prude : Dieu donne les enfants. Augustine était petite, ou, pour la mieux peindre, mignonne. Gracieuse et pleine de candeur, un homme du monde n’aurait pu reprocher à cette charmante créature que des gestes mesquins ou certaines attitudes communes, et parfois de la gêne. Sa figure silencieuse et immobile respirait cette mélancolie passagère qui s’empare de toutes les jeunes filles trop faibles pour oser résister aux volontés d’une mère. Toujours modestement vêtues, les deux soeurs ne pouvaient satisfaire la coquetterie innée chez la femme que par un luxe de propreté qui leur allait à merveille et les mettait en harmonie avec ces comptoirs luisants, avec ces rayons sur lesquels le vieux domestique ne souffrait pas un grain de poussière, avec la simplicité antique de tout ce qui se voyait autour d’elles. Obligées par leur genre de vie à chercher des éléments de bonheur dans des travaux obstinés, Augustine et Virginie n’avaient donné jusqu’alors que du contentement à leur mère, qui s’applaudissait secrètement de la perfection du caractère de ses deux filles. Il est facile d’imaginer les résultats de l’éducation qu’elles avaient reçue. Elevées pour le commerce, habituées à n’entendre que des raisonnements et des calculs tristement mercantiles, n’ayant étudié que la grammaire, la tenue des livres, un peu d’histoire juive, l’histoire de France dans Le Ragois, et ne lisant que les auteurs dont la lecture leur était permise par leur mère, leurs idées n’avaient pas pris beaucoup d’étendue : elles savaient parfaitement tenir un ménage, elles connaissaient le prix des choses, elles appréciaient les difficultés que l’on éprouve à amasser l’argent, elles étaient économes et portaient un grand respect aux qualités du négociant. Malgré la fortune de leur père, elles étaient aussi habiles à faire des reprises qu’à festonner ; souvent leur mère parlait de leur apprendre la cuisine afin qu’elles sussent bien ordonner un dîner, et pussent gronder une cuisinière en connaissance de cause. Ignorant les plaisirs du monde et voyant comment s’écoulait la vie exemplaire de leurs parents, elles ne jetaient que bien rarement leurs regards au delà de l’enceinte de cette vieille maison patrimoniale qui, pour leur mère, était l’univers. Les réunions occasionnées par les solennités de famille formaient tout l’avenir de leurs joies terrestres. Quand le grand salon situé au second étage devait recevoir madame Roguin, une demoiselle Chevrel, de quinze ans moins âgée que sa cousine et qui portait des diamants ; le jeune Rabourdin, sous-chef aux Finances ; monsieur César Birotteau, riche parfumeur, et sa femme appelée madame César ; monsieur Camusot, le plus riche négociant en soieries de la rue des Bourdonnais ; deux ou trois vieux banquiers, et des femmes irréprochables ; les apprêts nécessités par la manière dont l’argenterie, les porcelaines de Saxe, les bougies, les cristaux étaient empaquetés faisaient une diversion à la vie monotone de ces trois femmes qui allaient et venaient, en se donnant autant de mouvement que des religieuses pour la réception d’un évêque. Puis quand, le soir, fatiguées toutes trois d’avoir essuyé, frotté, déballé, mis en place les ornements de la fête, les deux jeunes filles aidaient leur mère à se coucher, madame Guillaume leur disait :\\n\\n— Nous n’avons rien fait aujourd’hui, mes enfants !\\n\\nLorsque, dans ces assemblées solennelles, la soeur tourière permettait de danser en confinant les parties de boston, de whist et de trictrac dans sa chambre à coucher, cette concession était comptée parmi les félicités les plus inespérées, et causait un bonheur égal à celui d’aller à deux ou trois grands bals où Guillaume menait ses filles à l’époque du carnaval. Enfin, une fois par an, l’honnête drapier donnait une fête pour laquelle rien n’était épargné. Quelque riches et élégantes que fussent les personnes invitées, elles se gardaient bien d’y manquer ; car les maisons les plus considérables de la place avaient recours à l’immense crédit, à la fortune ou à la vieille expérience de monsieur Guillaume. Mais les deux filles de ce digne négociant ne profitaient pas autant qu’on pourrait le supposer des enseignements que le monde offre à de jeunes âmes. Elles apportaient dans ces réunions, inscrites d’ailleurs sur le carnet d’échéances de la maison, des parures dont la mesquinerie les faisait rougir. Leur manière de danser n’avait rien de remarquable, et la surveillance maternelle ne leur permettait pas de soutenir la conversation autrement que par Oui et Non avec leurs cavaliers. Puis la loi de la vieille enseigne du Chat-qui-pelote leur ordonnait d’être rentrées à onze heures, moment où les bals et les fêtes commencent à s’animer. Ainsi leurs plaisirs, en apparence assez conformes à la fortune de leur père, devenaient souvent insipides par des circonstances qui tenaient aux habitudes et aux principes de cette famille. Quant à leur vie habituelle, une seule observation achèvera de la peindre. Madame Guillaume exigeait que ses deux filles fussent habillées de grand matin, qu’elles descendissent tous les jours à la même heure, et soumettait leurs occupations à une régularité monastique. Cependant Augustine avait reçu du hasard une âme assez élevée pour sentir le vide de cette existence. Parfois ses yeux bleus se relevaient comme pour interroger les profondeurs de cet escalier sombre et de ces magasins humides. Après avoir sondé ce silence de cloître, elle semblait écouter de loin de confuses révélations de cette vie passionnée qui met les sentiments à un plus haut prix que les choses. En ces moments son visage se colorait, ses mains inactives laissaient tomber la blanche mousseline sur le chêne poli du comptoir, et bientôt sa mère lui disait d’une voix qui restait toujours aigre même dans les tons les plus doux :\\n\\n— Augustine ! à quoi pensez-vous donc, mon bijou ?\\n\\nPeut-être Hippolyte comte de Douglas et le Comte de Comminges, deux romans trouvés par Augustine dans l’armoire d’une cuisinière récemment renvoyée par madame Guillaume, contribuèrent-ils à développer les idées de cette jeune fille qui les avait furtivement dévorés pendant les longues nuits de l’hiver précédent. Les expressions de désir vague, la voix douce, la peau de jasmin et les yeux bleus d’Augustine avaient donc allumé dans l’âme du pauvre Lebas un amour aussi violent que respectueux. Par un caprice facile à comprendre, Augustine ne se sentait aucun goût pour l’orphelin : peut-être était-ce parce qu’elle ne se savait pas aimée. En revanche, les longues jambes, les cheveux châtains, les grosses mains et l’encolure vigoureuse du premier commis avaient trouvé une secrète admiratrice dans mademoiselle Virginie, qui, malgré ses cinquante mille écus de dot, n’était demandée en mariage par personne. Rien de plus naturel que ces deux passions inverses nées dans le silence de ces comptoirs obscurs comme fleurissent des violettes dans la profondeur d’un bois. La muette et constante contemplation qui réunissait les yeux de ces jeunes gens par un besoin violent de distraction au milieu de travaux obstinés et d’une paix religieuse, devait tôt ou tard exciter des sentiments d’amour. L’habitude de voir une figure y fait découvrir insensiblement les qualités de l’âme, et finit par en effacer les défauts.\\n\\n— Au train dont y va cet homme, nos filles ne tarderont pas à se mettre à genoux devant un prétendu ! se dit monsieur Guillaume en lisant le premier décret par lequel Napoléon anticipa sur les classes de conscrits.\\n\\nDès ce jour, désespéré de voir sa fille aînée se faner, le vieux marchand se souvint d’avoir épousé mademoiselle Chevrel à peu près dans la situation où se trouvaient Joseph Lebas et Virginie. Quelle belle affaire que de marier sa fille et d’acquitter une dette sacrée, en rendant à un orphelin le bienfait qu’il avait reçu jadis de son prédécesseur dans les mêmes circonstances ! Agé de trente-trois ans, Joseph Lebas pensait aux obstacles que quinze ans de différence mettaient entre Augustine et lui. Trop perspicace d’ailleurs pour ne pas deviner les desseins de monsieur Guillaume, il en connaissait assez les principes inexorables pour savoir que jamais la cadette ne se marierait avant l’aînée. Le pauvre commis, dont le coeur était aussi excellent que ses jambes étaient longues et son buste épais, souffrait donc en silence.\\n\\nTel était l’état des choses dans cette petite république, qui, au milieu de la rue Saint-Denis, ressemblait assez à une succursale de la Trappe. Mais pour rendre un compte exact des événements extérieurs comme des sentiments, il est nécessaire de remonter à quelques mois avant la scène par laquelle commence cette histoire. A la nuit tombante, un jeune homme passant devant l’obscure boutique du Chat-qui-pelote y était resté un moment en contemplation à l’aspect d’un tableau qui aurait arrêté tous les peintres du monde. Le magasin, n’étant pas encore éclairé, formait un plan noir au fond duquel se voyait la salle à manger du marchand. Une lampe astrale y répandait ce jour jaune qui donne tant de grâce aux tableaux de l’école hollandaise. Le linge blanc, l’argenterie, les cristaux formaient de brillants accessoires qu’embellissaient encore de vives oppositions entre l’ombre et la lumière. La figure du père de famille et celle de sa femme, les visages des commis et les formes pures d’Augustine, à deux pas de laquelle se tenait une grosse fille joufflue, composaient un groupe si curieux ; ces têtes étaient si originales, et chaque caractère avait une expression si franche ; on devinait si bien la paix, le silence et la modeste vie de cette famille, que, pour un artiste accoutumé à exprimer la nature, il y avait quelque chose de désespérant à vouloir rendre cette scène fortuite. Ce passant était un jeune peintre, qui, sept ans auparavant, avait remporté le grand prix de peinture. Il revenait de Rome. Son âme nourrie de poésie, ses yeux rassasiés de Raphaël et de Michel-Ange, avaient soif de la nature vraie, après une longue habitation du pays pompeux où l’art a jeté partout son grandiose. Faux ou juste, tel était son sentiment personnel. Abandonné long-temps à la fougue des passions italiennes, son coeur demandait une de ces vierges modestes et recueillies que, malheureusement, il n’avait su trouver qu’en peinture à Rome. De l’enthousiasme imprimé à son âme exaltée par le tableau naturel qu’il contemplait, il passa naturellement à une profonde admiration pour la figure principale : Augustine paraissait pensive et ne mangeait point ; par une disposition de la lampe dont la lumière tombait entièrement sur son visage, son buste semblait se mouvoir dans un cercle de feu qui détachait plus vivement les contours de sa tête et l’illuminait d’une manière quasi surnaturelle. L’artiste la compara involontairement à un ange exilé qui se souvient du ciel. Une sensation presque inconnue, un amour limpide et bouillonnant inonda son coeur. Après être demeuré pendant un moment comme écrasé sous le poids de ses idées, il s’arracha à son bonheur, rentra chez lui, ne mangea pas, ne dormit point. Le lendemain, il entra dans son atelier pour n’en sortir qu’après avoir déposé sur une toile la magie de cette scène dont le souvenir l’avait en quelque sorte fanatisé. Sa félicité fut incomplète tant qu’il ne posséda pas un fidèle portrait de son idole. Il passa plusieurs fois devant la maison du Chat-qui-pelote ; il osa même y entrer une ou deux fois sous le masque d’un déguisement, afin de voir de plus près la ravissante créature que madame Guillaume couvrait de son aile. Pendant huit mois entiers, adonné à son amour, à ses pinceaux, il resta invisible pour ses amis les plus intimes, oubliant le monde, la poésie, le théâtre, la musique, et ses plus chères habitudes. Un matin, Girodet força toutes ces consignes que les artistes connaissent et savent éluder, parvint à lui et le réveilla par cette demande :\\n\\n— Que mettras-tu au Salon ?\\n\\nL’artiste saisit la main de son ami, l’entraîne à son atelier, découvre un petit tableau de chevalet et un portrait. Après une lente et avide contemplation des deux chefs-d’oeuvre, Girodet saute au cou de son camarade et l’embrasse, sans trouver de paroles. Ses émotions ne pouvaient se rendre que comme il les sentait, d’âme à âme.\\n\\n— Tu es amoureux ? dit Girodet.\\n\\nTous deux savaient que les plus beaux portraits de Titien, de Raphaël et de Léonard de Vinci sont dus à des sentiments exaltés, qui, sous diverses conditions, engendrent d’ailleurs tous les chefs-d’oeuvre. Pour toute réponse, le jeune artiste inclina la tête.\\n\\n— Es-tu heureux de pouvoir être amoureux ici, en revenant d’Italie ! Je ne te conseille pas de mettre de telles oeuvres au Salon, ajouta le grand peintre. Vois-tu, ces deux tableaux n’y seraient pas sentis. Ces couleurs vraies, ce travail prodigieux ne peuvent pas encore être appréciés, le public n’est plus accoutumé à tant de profondeur. Les tableaux que nous peignons, mon bon ami, sont des écrans, des paravents. Tiens, faisons plutôt des vers, et traduisons les Anciens ! il y a plus de gloire à en attendre, que de nos malheureuses toiles.\\n\\nMalgré cet avis charitable, les deux toiles furent exposées. La scène d’intérieur fit une révolution dans la peinture. Elle donna naissance à ces tableaux de genre dont la prodigieuse quantité importée à toutes nos expositions, pourrait faire croire qu’ils s’obtiennent par des procédés purement mécaniques. Quant au portrait, il est peu d’artistes qui ne gardent le souvenir de cette toile vivante à laquelle le public, quelquefois juste en masse, laissa la couronne que Girodet y plaça lui-même. Les deux tableaux furent entourés d’une foule immense. On s’y tua, comme disent les femmes. Des spéculateurs, des grands seigneurs couvrirent ces deux toiles de doubles napoléons, l’artiste refusa obstinément de les vendre, et refusa d’en faire des copies. On lui offrit une somme énorme pour les laisser graver, les marchands ne furent pas plus heureux que ne l’avaient été les amateurs. Quoique cette aventure fît du bruit dans le monde, elle n’était pas de nature à parvenir au fond de la petite Thébaïde de la rue Saint-Denis. Néanmoins, en venant faire une visite à madame Guillaume, la femme du notaire parla de l’exposition devant Augustine, qu’elle aimait beaucoup, et lui en expliqua le but. Le babil de madame Roguin inspira naturellement à Augustine le désir de voir les tableaux, et la hardiesse de demander secrètement à sa cousine de l’accompagner au Louvre. La cousine réussit dans la négociation qu’elle entama auprès de madame Guillaume, pour obtenir la permission d’arracher sa petite cousine à ses tristes travaux pendant environ deux heures. La jeune fille pénétra donc, à travers la foule, jusqu’au tableau couronné. Un frisson la fit trembler comme une feuille de bouleau, quand elle se reconnut. Elle eut peur et regarda autour d’elle pour rejoindre madame Roguin, de qui elle avait été séparée par un flot de monde. En ce moment ses yeux effrayés rencontrèrent la figure enflammée du jeune peintre. Elle se rappela tout à coup la physionomie d’un promeneur que, curieuse, elle avait souvent remarqué, en croyant que c’était un nouveau voisin.\\n\\n— Vous voyez ce que l’amour m’a fait faire, dit l’artiste à l’oreille de la timide créature qui resta tout épouvantée de ces paroles.\\n\\nElle trouva un courage surnaturel pour fendre la presse, et pour rejoindre sa cousine encore occupée à percer la masse du monde qui l’empêchait d’arriver jusqu’au tableau.\\n\\n— Vous seriez étouffée, s’écria Augustine, partons !\\n\\nMais il se rencontre, au Salon, certains moments pendant lesquels deux femmes ne sont pas toujours libres de diriger leurs pas dans les galeries. Mademoiselle Guillaume et sa cousine furent poussées à quelques pas du second tableau, par suite des mouvements irréguliers que la foule leur imprima. Le hasard voulut qu’elles eussent la facilité d’approcher ensemble de la toile illustrée par la mode, d’accord cette fois avec le talent. La femme du notaire fit une exclamation de surprise perdue dans le brouhaha et les bourdonnements de la foule ; mais Augustine pleura involontairement à l’aspect de cette merveilleuse scène. Puis, par un sentiment presque inexplicable, elle mit un doigt sur ses lèvres en apercevant à deux pas d’elle la figure extatique du jeune artiste. L’inconnu répondit par un signe de tête et désigna madame Roguin, comme un trouble-fête, afin de montrer à Augustine qu’elle était comprise. Cette pantomime jeta comme un brasier dans le corps de la pauvre fille qui se trouva criminelle, en se figurant qu’il venait de se conclure un pacte entre elle et l’artiste. Une chaleur étouffante, le continuel aspect des plus brillantes toilettes, et l’étourdissement que produisaient sur Augustine la variété des couleurs, la multitude des figures vivantes ou peintes, la profusion des cadres d’or, lui firent éprouver une espèce d’enivrement qui redoubla ses craintes. Elle se serait peut-être évanouie, si, malgré ce chaos de sensations, il ne s’était élevé au fond de son coeur une jouissance inconnue qui vivifia tout son être. Néanmoins, elle se crut sous l’empire de ce démon dont les terribles piéges lui étaient prédits par la voix tonnante des prédicateurs. Ce moment fut pour elle comme un moment de folie. Elle se vit accompagnée jusqu’à la voiture de sa cousine par ce jeune homme resplendissant de bonheur et d’amour. En proie à une irritation toute nouvelle, une ivresse qui la livrait en quelque sorte à la nature, Augustine écouta la voix éloquente de son coeur, et regarda plusieurs fois le jeune peintre en laissant paraître le trouble dont elle était saisie. Jamais l’incarnat de ses joues n’avait formé de plus vigoureux contrastes avec la blancheur de sa peau. L’artiste aperçut alors cette beauté dans toute sa fleur, cette pudeur dans toute sa gloire. Augustine éprouva une sorte de joie mêlée de terreur, en pensant que sa présence causait la félicité de celui dont le nom était sur toutes les lèvres, dont le talent donnait l’immortalité à de passagères images. Elle était aimée ! il lui était impossible d’en douter. Quand elle ne vit plus l’artiste, elle entendit encore retentir dans son coeur ces paroles simples : « Vous voyez ce que l’amour m’a fait faire. » Et les palpitations devenues plus profondes lui semblèrent une douleur, tant son sang plus ardent réveilla dans son corps de puissances inconnues. Elle feignit d’avoir un grand mal de tête pour éviter de répondre aux questions de sa cousine relativement aux tableaux ; mais, au retour, madame Roguin ne put s’empêcher de parler à madame Guillaume de la célébrité obtenue par le Chat-qui-pelote, et Augustine trembla de tous ses membres en entendant dire à sa mère qu’elle irait au Salon pour y voir sa maison. La jeune fille insista de nouveau sur sa souffrance, et obtint la permission d’aller se coucher.\\n\\n— Voilà ce qu’on gagne à tous ces spectacles, s’écria monsieur Guillaume, des maux de tête. Est-ce donc bien amusant de voir en peinture ce qu’on rencontre tous les jours dans notre rue ! Ne me parlez pas de ces artistes qui sont, comme vos auteurs, des meure-de-faim. Que diable ont-ils besoin de prendre ma maison pour la vilipender dans leurs tableaux ?\\n\\n— Cela pourra nous faire vendre quelques aunes de drap de plus, dit\\n\\nJoseph Lebas.\\n\\n\\n\\nCette observation n’empêcha pas que les arts et la pensée ne fussent condamnés encore une fois au tribunal du Négoce. Comme on doit bien le penser, ces discours ne donnèrent pas grand espoir à Augustine. Elle eut toute la nuit pour se livrer à la première méditation de l’amour. Les événements de cette journée furent comme un songe qu’elle se plut à reproduire dans sa pensée Elle s’initia aux craintes, aux espérances, aux remords, à toutes ces ondulations de sentiment qui devaient bercer un coeur simple et timide comme le sien. Quel vide elle reconnut dans cette noire maison, et quel trésor elle trouva dans son âme ! Etre la femme d’un homme de talent, partager sa gloire ! Quels ravages cette idée ne devait-elle pas faire au coeur d’une enfant élevée au sein de cette famille ! Quelle espérance ne devait-elle pas éveiller chez une jeune personne qui, nourrie jusqu’alors de principes vulgaires, avait désiré une vie élégante ! Un rayon de soleil était tombé dans cette prison. Augustine aima tout à coup. En elle tant de sentiments étaient flattés à la fois, qu’elle succomba sans rien calculer. A dix-huit ans, l’amour ne jette-t-il pas son prisme entre le monde et les yeux d’une jeune fille ? Incapable de deviner les rudes chocs qui résultent de l’alliance d’une femme aimante avec un homme d’imagination, elle crut être appelée à faire le bonheur de celui-ci, sans apercevoir aucune disparate entre elle et lui. Pour elle, le présent fut tout l’avenir. Quand le lendemain son père et sa mère revinrent du Salon, leurs figures attristées annoncèrent quelque désappointement. D’abord, les deux tableaux avaient été retirés par le peintre ; puis, madame Guillaume avait perdu son châle de cachemire. Apprendre que les tableaux venaient de disparaître après sa visite au Salon fut pour Augustine la révélation d’une délicatesse de sentiment que les femmes savent toujours apprécier, même instinctivement.\\n\\nLe matin où, rentrant d’un bal, Théodore de Sommervieux, tel était le nom que la renommée avait apporté dans le coeur d’Augustine, fut aspergé par les commis du Chat-qui-pelote pendant qu’il attendait l’apparition de sa naïve amie, qui ne le savait certes pas là , les deux amants se voyaient pour la quatrième fois seulement depuis la scène du Salon. Les obstacles que le régime de la maison Guillaume opposait au caractère fougueux de l’artiste, donnaient à sa passion pour Augustine une violence facile à concevoir. Comment aborder une jeune fille assise dans un comptoir entre deux femmes telles que mademoiselle Virginie et madame Guillaume ? Comment correspondre avec elle, quand sa mère ne la quittait jamais ? Habile, comme tous les amants, à se forger des malheurs, Théodore se créait un rival dans l’un des commis, et mettait les autres dans les intérêts de son rival. S’il échappait à tant d’Argus, il se voyait échouant sous les yeux sévères du vieux négociant ou de madame Guillaume. Partout des barrières, partout le désespoir ! La violence même de sa passion empêchait le jeune peintre de trouver ces expédients ingénieux qui, chez les prisonniers comme chez les amants, semblent être le dernier effort de la raison échauffée par un sauvage besoin de liberté ou par le feu de l’amour. Théodore tournait alors dans le quartier avec l’activité d’un fou, comme si le mouvement pouvait lui suggérer des ruses. Après s’être bien tourmenté l’imagination, il inventa de gagner à prix d’or la servante joufflue. Quelques lettres furent donc échangées de loin en loin pendant la quinzaine qui suivit la malencontreuse matinée où monsieur Guillaume et Théodore s’étaient si bien examinés.\\n\\nEn ce moment, les deux jeunes gens étaient convenus de se voir à une certaine heure du jour et le dimanche, à Saint-Leu, pendant la messe et les vêpres. Augustine avait envoyé à son cher Théodore la liste des parents et des amis de la famille, chez lesquels le jeune peintre tâcha d’avoir accès afin d’intéresser à ses amoureuses pensées, s’il était possible, une de ces âmes occupées d’argent, de commerce, et auxquelles une passion véritable devait sembler la spéculation la plus monstrueuse, une spéculation inouïe. D’ailleurs, rien ne changea dans les habitudes du Chat-qui-pelote. Si Augustine fut distraite, si, contre toute espèce d’obéissance aux lois de la charte domestique, elle monta à sa chambre pour y aller, grâce à un pot de fleurs, établir des signaux ; si elle soupira, si elle pensa enfin, personne, pas même sa mère, ne s’en aperçut. Cette circonstance causera quelque surprise à ceux qui auront compris l’esprit de cette maison, où une pensée entachée de poésie devait produire un contraste avec les êtres et les choses, où personne ne pouvait se permettre ni un geste, ni un regard qui ne fussent vus et analysés. Cependant rien de plus naturel : le vaisseau si tranquille qui naviguait sur la mer orageuse de la place de Paris, sous le pavillon du Chat-qui-pelote, était la proie d’une de ces tempêtes qu’on pourrait nommer équinoxiales à cause de leur retour périodique. Depuis quinze jours, les quatre hommes de l’équipage, madame Guillaume et mademoiselle Virginie s’adonnaient à ce travail excessif désigné sous le nom d’inventaire. On remuait tous les ballots et l’on vérifiait l’aunage des pièces pour s’assurer de la valeur exacte du coupon. On examinait soigneusement la carte appendue au paquet pour reconnaître en quel temps les draps avaient été achetés. On fixait le prix actuel. Toujours debout, son aune à la main, la plume derrière l’oreille, monsieur Guillaume ressemblait à un capitaine commandant la manoeuvre. Sa voix aiguë, passant par un judas pour interroger la profondeur des écoutilles du magasin d’en bas, faisait entendre ces barbares locutions du commerce, qui ne s’exprime que par énigmes : — Combien d’H-N-Z ? — Enlevé. — Que reste-t-il de Q-X ? — Deux aunes. — Quel prix ? — Cinq-cinq-trois. — Portez à trois A tout J-J, tout M-P, et le reste de V-D-O. Mille autres phrases tout aussi intelligibles ronflaient à travers les comptoirs comme des vers de la poésie moderne que des romantiques se seraient cités afin d’entretenir leur enthousiasme pour un de leurs poètes. Le soir, Guillaume, enfermé avec son commis et sa femme, soldait les comptes, portait à nouveau, écrivait aux retardataires, et dressait des factures. Tous trois préparaient ce travail immense dont le résultat tenait sur un carré de papier tellière, et prouvait à la maison Guillaume qu’il existait tant en argent, tant en marchandises, tant en traites et billets ; qu’elle ne devait pas un sou, qu’il lui était dû cent ou deux cent mille francs ; que le capital avait augmenté ; que les fermes, les maisons, les rentes allaient être ou arrondies, ou réparées, ou doublées. De là résultait la nécessité de recommencer avec plus d’ardeur que jamais à ramasser de nouveaux écus, sans qu’il vînt en tête à ces courageuses fourmis de se demander : A quoi bon ?\\n\\nA la faveur de ce tumulte annuel, l’heureuse Augustine échappait à l’investigation de ses Argus. Enfin, un samedi soir, la clôture de l’inventaire eut lieu. Les chiffres du total actif offrirent assez de zéros pour qu’en cette circonstance Guillaume levât la consigne sévère qui régnait toute l’année au dessert. Le sournois drapier se frotta les mains, et permit à ses commis de rester à sa table. A peine chacun des hommes de l’équipage achevait-il son petit verre d’une liqueur de ménage, on entendit le roulement d’une voiture. La famille alla voir Cendrillon aux Variétés, tandis que les deux derniers commis reçurent chacun un écu de six francs et la permission d’aller où bon leur semblerait, pourvu qu’ils fussent rentrés à minuit. Malgré cette débauche, le dimanche matin, le vieux marchand drapier fit sa barbe dès six heures, endossa son habit marron dont les superbes reflets lui causaient toujours le même contentement, il attacha des boucles d’or aux oreilles de son ample culotte de soie ; puis, vers sept heures, au moment où tout dormait encore dans la maison, il se dirigea vers le petit cabinet attenant à son magasin du premier étage. Le jour y venait d’une croisée armée de gros barreaux de fer, et qui donnait sur une petite cour carrée formée de murs si noirs qu’elle ressemblait assez à un puits. Le vieux négociant ouvrit lui-même ces volets garnis de tôle qu’il connaissait si bien, et releva une moitié du vitrage en le faisant glisser dans sa coulisse. L’air glacé de la cour vint rafraîchir la chaude atmosphère de ce cabinet, qui exhalait l’odeur particulière aux bureaux. Le marchand resta debout la main posée sur le bras crasseux d’un fauteuil de canne doublé de maroquin dont la couleur primitive était effacée, il semblait hésiter à s’y asseoir. Il regarda d’un air attendri le bureau à double pupitre, où la place de sa femme se trouvait ménagée, dans le côté opposé à la sienne, par une petite arcade pratiquée dans le mur. Il contempla les cartons numérotés, les ficelles, les ustensiles, les fers à marquer le drap, la caisse, objets d’une origine immémoriale, et crut se revoir devant l’ombre évoquée du sieur Chevrel. Il avança le même tabouret sur lequel il s’était jadis assis en présence de son défunt patron. Ce tabouret garni de cuir noir, et dont le crin s’échappait depuis long-temps par les coins mais sans se perdre, il le plaça d’une main tremblante au même endroit où son prédécesseur l’avait mis ; puis, dans une agitation difficile à décrire, il tira la sonnette qui correspondait au chevet du lit de Joseph Lebas. Quand ce coup décisif eut été frappé, le vieillard, pour qui ces souvenirs furent sans doute trop lourds, prit trois ou quatre lettres de change qui lui avaient été présentées, et les regarda sans les voir, quand Joseph Lebas se montra soudain.\\n\\n— Asseyez-vous là , lui dit Guillaume en lui désignant le tabouret.\\n\\nComme jamais le vieux maître-drapier n’avait fait asseoir son commis devant lui, Joseph Lebas tressaillit.\\n\\n— Que pensez-vous de ces traites ? demanda Guillaume.\\n\\n— Elles ne seront pas payées.\\n\\n— Comment ?\\n\\n— Mais j’ai su qu’avant-hier Etienne et compagnie ont fait leurs paiements en or.\\n\\n— Oh ! oh ! s’écria le drapier, il faut être bien malade pour laisser voir sa bile. Parlons d’autre chose. Joseph, l’inventaire est fini.\\n\\n— Oui, monsieur, et le dividende est un des plus beaux que vous ayez eus.\\n\\n— Ne vous servez donc pas de ces nouveaux mots ! Dites le produit, Joseph. Savez-vous, mon garçon, que c’est un peu à vous que nous devons ces résultats ! aussi, ne veux-je plus que vous ayez d’appointements. Madame Guillaume m’a donné l’idée de vous offrir un intérêt. Hein, Joseph ! Guillaume et Lebas, ces mots ne feraient-ils pas une belle raison sociale ? On pourrait mettre et compagnie pour arrondir la signature.\\n\\nLes larmes vinrent aux yeux de Joseph Lebas, qui s’efforça de les cacher.\\n\\n— Ah, monsieur Guillaume ! comment ai-je pu mériter tant de bontés ? Je n’ai fait que mon devoir. C’était déjà tant que de vous intéresser à un pauvre orph…\\n\\nIl brossait le parement de sa manche gauche avec la manche droite, et n’osait regarder le vieillard qui souriait en pensant que ce modeste jeune homme avait sans doute besoin, comme lui autrefois, d’être encouragé pour rendre l’explication complète.\\n\\n— Cependant, reprit le père de Virginie, vous ne méritez pas beaucoup cette faveur, Joseph ! Vous ne mettez pas en moi autant de confiance que j’en mets en vous. (Le commis releva brusquement la tête.)\\n\\n— Vous avez le secret de la caisse. Depuis deux ans je vous ai dit presque toutes mes affaires. Je vous ai fait voyager en fabrique. Enfin, pour vous, je n’ai rien sur le coeur. Mais vous ?… vous avez une inclination, et ne m’en avez pas touché un seul mot. (Joseph Lebas rougit.)\\n\\n— Ah ! ah ! s’écria Guillaume, vous pensiez donc tromper un vieux renard comme moi ? Moi ! à qui vous avez vu deviner la faillite Lecoq.\\n\\n— Comment, monsieur ? répondit Joseph Lebas en examinant son patron avec autant d’attention que son patron l’examinait, comment, vous sauriez qui j’aime ?\\n\\n— Je sais tout, vaurien, lui dit le respectable et rusé marchand en lui tordant le bout de l’oreille. Et je te pardonne, j’ai fait de même.\\n\\n— Et vous me l’accorderiez ?\\n\\n— Oui, avec cinquante mille écus, et je t’en laisserai autant, et nous marcherons sur nouveaux frais avec une nouvelle raison sociale. Nous brasserons encore des affaires, garçon, s’écria le vieux marchand en s’exaltant, se levant et agitant ses bras. Vois-tu, mon gendre, il n’y a que le commerce ! Ceux qui se demandent quels plaisirs on y trouve sont des imbéciles. Etre à la piste des affaires, savoir gouverner sur la place, attendre avec anxiété, comme au jeu, si les Etienne et compagnie font faillite, voir passer un régiment de la garde impériale habillé de notre drap, donner un croc en jambe au voisin, loyalement s’entend ! fabriquer à meilleur marché que les autres ; suivre une affaire qu’on ébauche, qui commence, grandit, chancelle et réussit ; connaître comme un ministre de la police tous les ressorts des maisons de commerce pour ne pas faire fausse route ; se tenir debout devant les naufrages ; avoir des amis, par correspondance, dans toutes les villes manufacturières, n’est-ce pas un jeu perpétuel, Joseph ? Mais c’est vivre, ça ! Je mourrai dans ce tracas-là , comme le vieux Chevrel, n’en prenant cependant plus qu’à mon aise. Dans la chaleur de sa plus forte improvisation, le père Guillaume n’avait presque pas regardé son commis qui pleurait à chaudes larmes.\\n\\n— Eh bien ! Joseph, mon pauvre garçon, qu’as-tu donc ?\\n\\n— Ah ! je l’aime tant, tant, monsieur Guillaume, que le coeur me manque, je crois…\\n\\n— Eh bien ! garçon, dit le marchand attendri, tu es plus heureux que tu ne crois, sarpejeu, car elle t’aime. Je le sais, moi !\\n\\nEt il cligna ses deux petits yeux verts en regardant son commis.\\n\\n— Mademoiselle Augustine, mademoiselle Augustine ! s’écria Joseph\\n\\nLebas dans son enthousiasme.\\n\\n\\n\\nIl allait s’élancer hors du cabinet, quand il se sentit arrêté par un bras de fer, et son patron stupéfait le ramena vigoureusement devant lui.\\n\\n— Qu’est-ce que fait donc Augustine dans cette affaire-là ? demanda Guillaume dont la voix glaça sur-le-champ le malheureux Joseph Lebas.\\n\\n— N’est-ce pas elle… que… j’aime ? dit le commis en balbutiant. Déconcerté de son défaut de perspicacité, Guillaume se rassit et mit sa tête pointue dans ses deux mains pour réfléchir à la bizarre position dans laquelle il se trouvait. Joseph Lebas honteux et au désespoir resta debout.\\n\\n— Joseph, reprit le négociant avec une dignité froide, je vous parlais de Virginie. L’amour ne se commande pas, je le sais. Je connais votre discrétion, nous oublierons cela. Je ne marierai jamais Augustine avant Virginie. Votre intérêt sera de dix pour cent.\\n\\nLe commis, auquel l’amour donna je ne sais quel degré de courage et d’éloquence, joignit les mains, prit la parole, parla pendant un quart d’heure à Guillaume avec tant de chaleur et de sensibilité, que la situation changea. S’il s’était agi d’une affaire commerciale, le vieux négociant aurait eu des règles fixes pour prendre une résolution ; mais, jeté à mille lieues du commerce, sur la mer des sentiments, et sans boussole, il flotta irrésolu devant un événement si original, se disait-il. Entraîné par sa bonté naturelle, il battit un peu la campagne.\\n\\n— Et, diantre, Joseph, tu n’es pas sans savoir que j’ai eu mes deux enfants à dix ans de distance ! Mademoiselle Chevrel n’était pas belle, elle n’a cependant pas à se plaindre de moi. Fais donc comme moi. Enfin, ne pleure pas, es-tu bête ? Que veux-tu ? cela s’arrangera peut-être, nous verrons. Il y a toujours moyen de se tirer d’affaire. Nous autres hommes nous ne sommes pas toujours comme des Céladons pour nos femmes. Tu m’entends ? Madame Guillaume est dévote, et… Allons, sarpejeu, mon enfant, donne ce matin le bras à Augustine pour aller à la messe.\\n\\nTelles furent les phrases jetées à l’aventure par Guillaume. La conclusion qui les terminait ravit l’amoureux commis : il songeait déjà pour mademoiselle Virginie à l’un de ses amis, quand il sortit du cabinet enfumé en serrant la main de son futur beau-père, après lui avoir dit, d’un petit air entendu, que tout s’arrangerait au mieux. « Que va penser madame Guillaume ? » Cette idée tourmenta prodigieusement le brave négociant quand il fut seul.\\n\\nAu déjeuner, madame Guillaume et Virginie, auxquelles le marchand-drapier avait laissé provisoirement ignorer son désappointement, regardèrent assez malicieusement Joseph Lebas qui resta grandement embarrassé. La pudeur du commis lui concilia l’amitié de sa belle-mère. La matrone redevint si gaie qu’elle regarda monsieur Guillaume en souriant, et se permit quelques petites plaisanteries d’un usage immémorial dans ces innocentes familles. Elle mit en question la conformité de la taille de Virginie et de celle de Joseph, pour leur demander de se mesurer. Ces niaiseries préparatoires attirèrent quelques nuages sur le front du chef de famille, et il afficha même un tel amour pour le décorum, qu’il ordonna à Augustine de prendre le bras du premier commis en allant à Saint-Leu. Madame Guillaume, étonnée de cette délicatesse masculine, honora son mari d’un signe de tête d’approbation. Le cortége partit donc de la maison dans un ordre qui ne pouvait suggérer aucune interprétation malicieuse aux voisins.\\n\\n— Ne trouvez-vous pas, mademoiselle Augustine, disait le commis en tremblant, que la femme d’un négociant qui a un bon crédit, comme monsieur Guillaume, par exemple, pourrait s’amuser un peu plus que ne s’amuse madame votre mère, pourrait porter des diamants, aller en voiture ? Oh ! moi, d’abord, si je me mariais, je voudrais avoir toute la peine, et voir ma femme heureuse. Je ne la mettrais pas dans mon comptoir. Voyez-vous, dans la draperie, les femmes n’y sont plus aussi nécessaires qu’elles l’étaient autrefois. Monsieur Guillaume a eu raison d’agir comme il a fait, et d’ailleurs c’était le goût de son épouse. Mais qu’une femme sache donner un coup de main à la comptabilité, à la correspondance, au détail, aux commandes, à son ménage, afin de ne pas rester oisive, c’est tout. A sept heures, quand la boutique serait fermée, moi je m’amuserais, j’irais au spectacle et dans le monde. Mais vous ne m’écoutez pas.\\n\\n— Si fait, monsieur Joseph. Que dites-vous de la peinture ? C’est là un bel état.\\n\\n— Oui, je connais un maître peintre en bâtiment, monsieur Lourdois, qui a des écus.\\n\\nEn devisant ainsi, la famille atteignit l’église de Saint-Leu. Là , madame Guillaume retrouva ses droits, et fit mettre, pour la première fois, Augustine à côté d’elle. Virginie prit place sur la quatrième chaise à côté de Lebas. Pendant le prône, tout alla bien entre Augustine et Théodore qui, debout derrière un pilier, priait sa madone avec ferveur ; mais au lever-Dieu, madame Guillaume s’aperçut, un peu tard, que sa fille Augustine tenait son livre de messe au rebours. Elle se disposait à la gourmander vigoureusement, quand, rabaissant son voile, elle interrompit sa lecture et se mit à regarder dans la direction qu’affectionnaient les yeux de sa fille. A l’aide de ses bésicles, elle vit le jeune artiste dont l’élégance mondaine annonçait plutôt quelque capitaine de cavalerie en congé, qu’un négociant du quartier. Il est difficile d’imaginer l’état violent dans lequel se trouva madame Guillaume, qui se flattait d’avoir parfaitement élevé ses filles, en reconnaissant dans le coeur d’Augustine un amour clandestin dont le danger lui fut exagéré par sa pruderie et par son ignorance. Elle crut sa fille gangrenée jusqu’au coeur.\\n\\n— Tenez d’abord votre livre à l’endroit, mademoiselle, dit-elle à voix basse mais en tremblant de colère. Elle arracha vivement le Paroissien accusateur, et le remit de manière à ce que les lettres fussent dans leur sens naturel.\\n\\n— N’ayez pas le malheur de lever les yeux autre part que sur vos prières, ajouta-t-elle, autrement, vous auriez affaire à moi. Après la messe, votre père et moi nous aurons à vous parler.\\n\\nCes paroles furent comme un coup de foudre pour la pauvre Augustine. Elle se sentit défaillir ; mais combattue entre la douleur qu’elle éprouvait et la crainte de faire un esclandre dans l’église, elle eut le courage de cacher ses angoisses. Cependant, il était facile de deviner l’état violent de son âme en voyant son Paroissien trembler et des larmes tomber sur chacune des pages qu’elle tournait. Au regard enflammé que lui lança madame Guillaume, l’artiste vit le péril où tombaient ses amours, et sortit, la rage dans le coeur, décidé à tout oser.\\n\\n— Allez dans votre chambre, mademoiselle ! dit madame Guillaume à sa fille en rentrant au logis ; nous vous ferons appeler ; et surtout, ne vous avisez pas d’en sortir.\\n\\nLa conférence que les deux époux eurent ensemble fut si secrète, que rien n’en transpira d’abord. Cependant, Virginie, qui avait encouragé sa soeur par mille douces représentations, poussa la complaisance jusqu’à se glisser auprès de la porte de la chambre à coucher de sa mère, chez laquelle la discussion avait lieu, pour y recueillir quelques phrases. Au premier voyage qu’elle fit du troisième au second étage, elle entendit son père qui s’écriait :\\n\\n— Madame, vous voulez donc tuer votre fille ?\\n\\n— Ma pauvre enfant, dit Virginie à sa soeur éplorée, papa prend ta défense !\\n\\n— Et que veulent-ils faire à Théodore ? demanda l’innocente créature.\\n\\nLa curieuse Virginie redescendit alors ; mais cette fois elle resta plus long-temps : elle apprit que Lebas aimait Augustine. Il était écrit que, dans cette mémorable journée, une maison ordinairement si calme serait un enfer. Monsieur Guillaume désespéra Joseph Lebas en lui confiant l’amour d’Augustine pour un étranger. Lebas, qui avait averti son ami de demander mademoiselle Virginie en mariage, vit ses espérances renversées. Mademoiselle Virginie, accablée de savoir que Joseph l’avait en quelque sorte refusée, fut prise d’une migraine. La zizanie, semée entre les deux époux par l’explication que monsieur et madame Guillaume avaient eue ensemble, et où, pour la troisième fois de leur vie, ils se trouvèrent d’opinions différentes, se manifesta d’une manière terrible. Enfin, à quatre heures après midi, Augustine, pâle, tremblante et les yeux rouges, comparut devant son père et sa mère. La pauvre enfant raconta naïvement la trop courte histoire de ses amours. Rassurée par l’allocution de son père, qui lui avait promis de l’écouter en silence, elle prit un certain courage en prononçant devant ses parents le nom de son cher Théodore de Sommervieux, et en fit malicieusement sonner la particule aristocratique. En se livrant au charme inconnu de parler de ses sentiments, elle trouva assez de hardiesse pour déclarer avec une innocente fermeté qu’elle aimait monsieur de Sommervieux, qu’elle le lui avait écrit, et ajouta, les larmes aux yeux :\\n\\n— Ce serait faire mon malheur que de me sacrifier à un autre.\\n\\n— Mais, Augustine, vous ne savez donc pas ce que c’est qu’un peintre ? s’écria sa mère avec horreur.\\n\\n— Madame Guillaume ! dit le vieux père en imposant silence à sa femme.\\n\\n— Augustine, dit-il, les artistes sont en général des meure-de-faim. Ils sont trop dépensiers pour ne pas être toujours de mauvais sujets. J’ai fourni feu M. Joseph Vernet, feu M. Lekain et feu M. Noverre. Ah ! si tu savais combien ce M. Noverre, M. le chevalier de Saint-Georges, et surtout M. Philidor, ont joué de tours à ce pauvre père Chevrel ! Ce sont de drôles de corps, je le sais bien. Ça vous a tous un babil, des manières… Ah ! jamais ton monsieur Sumer… Somm…\\n\\n— De Sommervieux, mon père !\\n\\n— Eh bien ! de Sommervieux, soit ! Jamais il n’aura été aussi agréable avec toi que M. le chevalier de Saint-Georges le fut avec moi, le jour où j’obtins une sentence des consuls contre lui. Aussi était-ce des gens de qualité d’autrefois.\\n\\n— Mais, mon père, monsieur Théodore est noble, et m’a écrit qu’il était riche. Son père s’appelait le chevalier de Sommervieux avant la révolution.\\n\\nA ces paroles, monsieur Guillaume regarda sa terrible moitié, qui, en femme contrariée frappait le plancher du bout du pied et gardait un morne silence. Elle évitait même de jeter ses yeux courroucés sur Augustine, et semblait laisser à monsieur Guillaume toute la responsabilité d’une affaire si grave, puisque ses avis n’étaient pas écoutés. Cependant, malgré son flegme apparent, quand elle vit son mari prenant si doucement son parti sur une catastrophe qui n’avait rien de commercial, elle s’écria :\\n\\n\\n\\n\\n\\n— En vérité, monsieur, vous êtes d’une faiblesse avec vos filles… mais…\\n\\nLe bruit d’une voiture qui s’arrêtait à la porte interrompit tout à coup la mercuriale que le vieux négociant redoutait déjà . En un moment, madame Roguin se trouva au milieu de la chambre, et, regardant les trois acteurs de cette scène domestique :\\n\\n— Je sais tout, ma cousine, dit-elle d’un air de protection.\\n\\nMadame Roguin avait un défaut, celui de croire que la femme d’un notaire de Paris pouvait jouer le rôle d’une petite maîtresse.\\n\\n— Je sais tout, répéta-t-elle, et je viens dans l’arche de Noé, comme la colombe, avec la branche d’olivier. J’ai lu cette allégorie dans le Génie du christianisme, dit-elle en se retournant vers madame Guillaume, la comparaison doit vous plaire, ma cousine. Savez-vous, ajouta-t-elle en souriant à Augustine, que ce monsieur de Sommervieux est un homme charmant ? Il m’a donné ce matin mon portrait fait de main de maître. Cela vaut au moins six mille francs.\\n\\nA ces mots, elle frappa doucement sur les bras de monsieur Guillaume. Le vieux négociant ne put s’empêcher de faire avec ses lèvres une grosse moue qui lui était particulière.\\n\\n— Je connais beaucoup monsieur de Sommervieux, reprit la colombe. Depuis une quinzaine de jours il vient à mes soirées, il en fait le charme. Il m’a conté toutes ses peines et m’a prise pour avocat. Je sais de ce matin qu’il adore Augustine, et il l’aura. Ah ! cousine, n’agitez pas ainsi la tête en signe de refus. Apprenez qu’il sera créé baron, et qu’il vient d’être nommé chevalier de la Légion-d’Honneur par l’empereur lui-même, au Salon. Roguin est devenu son notaire et connaît ses affaires. Eh bien ! monsieur de Sommervieux possède en bons biens au soleil douze mille livres de rente. Savez-vous que le beau-père d’un homme comme lui peut devenir quelque chose, maire de son arrondissement, par exemple ! N’avez-vous pas vu monsieur Dupont être fait comte de l’empire et sénateur pour être venu, en sa qualité de maire, complimenter l’empereur sur son entrée à Vienne. Oh ! ce mariage-là se fera. Je l’adore, moi, ce bon jeune homme. Sa conduite envers Augustine ne se voit que dans les romans. Va, ma petite, tu seras heureuse, et tout le monde voudrait être à ta place. J’ai chez moi, à mes soirées, madame la duchesse de Carigliano qui raffole de monsieur de Sommervieux. Quelques méchantes langues disent qu’elle ne vient chez moi que pour lui, comme si une duchesse d’hier était déplacée chez une Chevrel dont la famille a cent ans de bonne bourgeoisie.\\n\\n— Augustine, reprit madame Roguin après une petite pause, j’ai vu le portrait. Dieu ! qu’il est beau. Sais-tu que l’empereur a voulu le voir ? Il a dit en riant au Vice-Connétable que s’il y avait beaucoup de femmes comme celle-là à sa cour pendant qu’il y venait tant de rois, il se faisait fort de maintenir toujours la paix en Europe. Est-ce flatteur ?\\n\\nLes orages par lesquels cette journée avait commencé devaient ressembler à ceux de la nature, en ramenant un temps calme et serein. Madame Roguin déploya tant de séductions dans ses discours, elle sut attaquer tant de cordes à la fois dans les coeurs secs de monsieur et de madame Guillaume, qu’elle finit par en trouver une dont elle tira parti. A cette singulière époque, le commerce et la finance avaient plus que jamais la folle manie de s’allier aux grands seigneurs, et les généraux de l’empire profitèrent assez bien de ces dispositions. Monsieur Guillaume s’élevait singulièrement contre cette déplorable passion. Ses axiomes favoris étaient que, pour trouver le bonheur, une femme devait épouser un homme de sa classe ; on était toujours tôt ou tard puni d’avoir voulu monter trop haut ; l’amour résistait si peu aux tracas du ménage, qu’il fallait trouver l’un chez l’autre des qualités bien solides pour être heureux ; il ne fallait pas que l’un des deux époux en sût plus que l’autre, parce qu’on devait avant tout se comprendre ; un mari qui parlait grec et la femme latin, risquaient de mourir de faim. Il avait inventé cette espèce de proverbe. Il comparait les mariages ainsi faits à ces anciennes étoffes de soie et de laine, dont la soie finissait toujours par couper la laine. Cependant, il se trouve tant de vanité au fond du coeur de l’homme, que la prudence du pilote qui gouvernait si bien le Chat-qui-pelote, succomba sous l’agressive volubilité de madame Roguin. La sévère madame Guillaume, la première, trouva dans l’inclination de sa fille des motifs pour déroger à ces principes, et pour consentir à recevoir au logis monsieur de Sommervieux, qu’elle se promit de soumettre à un rigoureux examen.\\n\\nLe vieux négociant alla trouver Joseph Lebas, et l’instruisit de l’état des choses. A six heures et demie, la salle à manger illustrée par le peintre, réunit sous son toit de verre, madame et monsieur Roguin, le jeune peintre et sa charmante Augustine, Joseph Lebas qui prenait son bonheur en patience, et mademoiselle Virginie dont la migraine avait cessé. Monsieur et madame Guillaume virent en perspective leurs enfants établis et les destinées du Chat-qui-pelote remises en des mains habiles. Leur contentement fut au comble, quand, au dessert, Théodore leur fit présent de l’étonnant tableau qu’ils n’avaient pu voir, et qui représentait l’intérieur de cette vieille boutique, à laquelle était dû tant de bonheur.\\n\\n— C’est-y gentil, s’écria Guillaume. Dire qu’on voulait donner trente mille francs de cela.\\n\\n— Mais c’est qu’on y trouve mes barbes, reprit madame Guillaume.\\n\\n— Et ces étoffes dépliées, ajouta Lebas, on les prendrait avec la main.\\n\\n— Les draperies font toujours très-bien, répondit le peintre. Nous serions trop heureux, nous autres artistes modernes, d’atteindre à la perfection de la draperie antique.\\n\\n— Vous aimez donc la draperie, s’écria le père Guillaume. Eh bien, sarpejeu ! touchez là , mon jeune ami. Puisque vous estimez le commerce, nous nous entendrons. Eh ! pourquoi le mépriserait-on ? Le monde a commencé par là , puisque Adam a vendu le paradis pour une pomme. Ça n’a pas été une fameuse spéculation, par exemple !\\n\\nEt le vieux négociant se mit à éclater d’un gros rire franc excité par le vin de Champagne qu’il faisait circuler généreusement. Le bandeau qui couvrait les yeux du jeune artiste fut si épais qu’il trouva ses futurs parents aimables. Il ne dédaigna pas de les égayer par quelques charges de bon goût. Aussi plut-il généralement. Le soir, quand le salon meublé de choses très-cossues, pour se servir de l’expression de Guillaume, fut désert ; pendant que madame Guillaume s’en allait de table en cheminée, de candélabre en flambeau, soufflant avec précipitation les bougies, le brave négociant, qui savait toujours voir clair aussitôt qu’il s’agissait d’affaires ou d’argent, attira sa fille Augustine auprès de lui ; puis, après l’avoir prise sur ses genoux, il lui tint ce discours :\\n\\n— Ma chère enfant, tu épouseras ton Sommervieux, puisque tu le veux ; permis à toi de risquer ton capital de bonheur. Mais je ne me laisse pas prendre à ces trente mille francs que l’on gagne à gâter de bonnes toiles. L’argent qui vient si vite s’en va de même. N’ai-je pas entendu dire ce soir à ce jeune écervelé que si l’argent était rond, c’était pour rouler ! S’il est rond pour les gens prodigues, il est plat pour les gens économes qui l’empilent et l’amassent. Or, mon enfant, ce beau garçon-là parle de te donner des voitures, des diamants ? Il a de l’argent, qu’il le dépense pour toi ! bene sit ! Je n’ai rien à y voir. Mais quant à ce que je te donne, je ne veux pas que des écus si péniblement ensachés s’en aillent en carrosses ou en colifichets. Qui dépense trop n’est jamais riche. Avec les cent mille écus de sa dot on n’achète pas encore tout Paris. Tu as beau avoir à recueillir un jour quelques centaines de mille francs, je te les ferai attendre, sarpejeu ! le plus long-temps possible. J’ai donc attiré ton prétendu dans un coin, et un homme qui a mené la faillite Lecocq n’a pas eu grande peine à faire consentir un artiste à se marier séparé de biens avec sa femme. J’aurai l’oeil au contrat pour bien faire stipuler les donations qu’il se propose de te constituer. Allons, mon enfant, j’espère être grand-père, sarpejeu ! je veux m’occuper déjà de mes petits-enfants : jure-moi donc ici de ne jamais rien signer en fait d’argent que par mon conseil ; et si j’allais trouver trop tôt le père Chevrel, jure-moi de consulter le jeune Lebas, ton beau-frère. Promets-le-moi.\\n\\n— Oui, mon père, je vous le jure.\\n\\nA ces mots prononcés d’une voix douce, le vieillard baisa sa fille sur les deux joues. Ce soir-là , tous les amants dormirent presque aussi paisiblement que monsieur et madame Guillaume. Quelques mois après ce mémorable dimanche, le maître-autel de Saint-Leu fut témoin de deux mariages bien différents. Augustine et Théodore s’y présentèrent dans tout l’éclat du bonheur, les yeux pleins d’amour, parés de toilettes élégantes, attendus par un brillant équipage. Venue dans un bon remise avec sa famille, Virginie, donnant le bras à son père, suivait sa jeune soeur humblement et dans de plus simples atours, comme une ombre nécessaire aux harmonies de ce tableau. Monsieur Guillaume s’était donné toutes les peines imaginables pour obtenir à l’église que Virginie fût mariée avant Augustine ; mais il eut la douleur de voir le haut et le bas clergé s’adresser en toute circonstance à la plus élégante des mariées. Il entendit quelques-uns de ses voisins approuver singulièrement le bon sens de mademoiselle Virginie, qui faisait, disaient-ils, le mariage le plus solide, et restait fidèle au quartier ; tandis qu’ils lancèrent quelques brocards suggérés par l’envie sur Augustine qui épousait un artiste, un noble ; ils ajoutèrent avec une sorte d’effroi que, si les Guillaume avaient de l’ambition, la draperie était perdue. Un vieux marchand d’éventails ayant dit que ce mange-tout-là l’aurait bientôt mise sur la paille, le père Guillaume s’applaudit in petto de la prudence qu’il avait mise dans la rédaction des conventions matrimoniales. Le soir, la famille se sépara après un bal somptueux, suivi d’un de ces soupers plantureux dont le souvenir commence à se perdre dans la génération présente. Monsieur et madame Guillaume restèrent dans leur hôtel de la rue du Colombier où la noce avait eu lieu. Monsieur et madame Lebas retournèrent dans leur remise à la vieille maison de la rue Saint-Denis pour y diriger la nauf du Chat-qui-pelote. L’artiste, ivre de bonheur, prit entre ses bras sa chère Augustine, l’enleva vivement quand leur coupé arriva rue des Trois-Frères, et la porta dans son élégant appartement.\\n\\nLa fougue de passion qui possédait Théodore fit dévorer au jeune ménage près d’une année entière sans que le moindre nuage vînt altérer l’azur du ciel sous lequel ils vivaient. Pour eux, l’existence n’eut rien de pesant. Théodore répandait sur chaque journée d’incroyables fioriture de plaisirs. Il se plaisait à varier les emportements de la passion, par la molle langueur de ces repos où les âmes sont lancées si haut dans l’extase qu’elles semblent y oublier l’union corporelle. Incapable de réfléchir, l’heureuse Augustine se prêtait à l’allure onduleuse de son bonheur. Elle ne croyait pas faire encore assez en se livrant toute à l’amour permis et saint du mariage. Simple et naïve, elle ne connaissait ni la coquetterie des refus, ni l’empire qu’une jeune demoiselle du grand monde se crée sur un mari par d’adroits caprices. Elle aimait trop pour calculer l’avenir, et n’imaginait pas qu’une vie si délicieuse pût jamais cesser. Heureuse d’être alors tous les plaisirs de son mari, elle crut que cet inextinguible amour serait toujours pour elle la plus belle de toutes les parures, comme son dévouement et son obéissance seraient un éternel attrait. Enfin, la félicité de l’amour l’avait rendue si brillante, que sa beauté lui inspira de l’orgueil et lui donna la conscience de pouvoir toujours régner sur un homme aussi facile à enflammer que monsieur de Sommervieux. Ainsi son état de femme ne lui apporta d’autres enseignements que ceux de l’amour. Au sein de ce bonheur, elle resta l’ignorante petite fille qui vivait obscurément rue Saint-Denis, et ne pensa point à prendre les manières, l’instruction, le ton du monde dans lequel elle devait vivre. Ses paroles étant des paroles d’amour, elle y déployait bien une sorte de souplesse d’esprit et une certaine délicatesse d’expression ; mais elle se servait du langage commun à toutes les femmes quand elles se trouvent plongées dans une passion qui semble être leur élément. Si, par hasard, une idée discordante avec celles de Théodore était exprimée par Augustine, le jeune artiste en riait comme on rit des premières fautes que fait un étranger, mais qui finissent par fatiguer s’il ne se corrige pas.\\n\\nCependant, à l’expiration de cette année aussi charmante que rapide, Sommervieux sentit un matin la nécessité de reprendre ses travaux et ses habitudes. Sa femme était enceinte. Il revit ses amis. Pendant les longues souffrances de l’année où, pour la première fois, une jeune femme nourrit un enfant, il travailla sans doute avec ardeur ; mais parfois il retourna chercher quelques distractions dans le grand monde. La maison où il allait le plus volontiers était celle de la duchesse de Carigliano qui avait fini par attirer chez elle le célèbre artiste. Quand Augustine fut rétablie, quand son fils ne réclama plus ces soins assidus qui interdisent à une mère les plaisirs du monde, Théodore en était arrivé à vouloir éprouver cette jouissance d’amour-propre que nous donne la société quand nous y apparaissons avec une belle femme, objet d’envie et d’admiration. Parcourir les salons en s’y montrant avec l’éclat emprunté de la gloire de son mari, se voir jalousée par toutes les femmes, fut pour Augustine une nouvelle moisson de plaisirs ; mais ce fut le dernier reflet que devait jeter son bonheur conjugal. Elle commença par offenser la vanité de son mari, quand, malgré de vains efforts, elle laissa percer son ignorance, l’impropriété de son langage et l’étroitesse de ses idées. Le caractère de Sommervieux, dompté pendant près de deux ans et demi par les premiers emportements de l’amour, reprit, avec la tranquillité d’une possession moins jeune, sa pente et ses habitudes un moment détournées de leur cours. La poésie, la peinture et les exquises jouissances de l’imagination possèdent sur les esprits élevés des droits imprescriptibles. Ces besoins d’une âme forte n’avaient pas été trompés chez Théodore pendant ces deux années, ils avaient trouvé seulement une pâture nouvelle. Quand les champs de l’amour furent parcourus, quand l’artiste eut, comme les enfants, cueilli des roses et des bleuets avec une telle avidité qu’il ne s’apercevait pas que ses mains ne pouvaient plus les tenir, la scène changea. Si le peintre montrait à sa femme les croquis de ses plus belles compositions, il l’entendait s’écrier comme eût fait le père Guillaume : « C’est bien joli ! » Son admiration sans chaleur ne provenait pas d’un sentiment consciencieux, mais de la croyance sur parole de l’amour. Augustine préférait un regard au plus beau tableau. Le seul sublime qu’elle connût était celui du coeur. Enfin, Théodore ne put se refuser à l’évidence d’une vérité cruelle : sa femme n’était pas sensible à la poésie, elle n’habitait pas sa sphère, elle ne le suivait pas dans tous ses caprices, dans ses improvisations, dans ses joies, dans ses douleurs ; elle marchait terre à terre dans le monde réel, tandis qu’il avait la tête dans les cieux. Les esprits ordinaires ne peuvent pas apprécier les souffrances renaissantes de l’être qui, uni à un autre par le plus intime de tous les sentiments, est obligé de refouler sans cesse les plus chères expansions de sa pensée, et de faire rentrer dans le néant les images qu’une puissance magique le force à créer. Pour lui, ce supplice est d’autant plus cruel, que le sentiment qu’il porte à son compagnon ordonne, par sa première loi, de ne jamais rien se dérober l’un à l’autre, et de confondre les effusions de la pensée aussi bien que les épanchements de l’âme. On ne trompe pas impunément les volontés de la nature : elle est inexorable comme la Nécessité, qui, certes, est une sorte de nature sociale. Sommervieux se réfugia dans le calme et le silence de son atelier, en espérant que l’habitude de vivre avec des artistes pourrait former sa femme, et développerait en elle les germes de haute intelligence engourdis que quelques esprits supérieurs croient préexistants chez tous les êtres ; mais Augustine était trop sincèrement religieuse pour ne pas être effrayée du ton des artistes. Au premier dîner que donna Théodore, elle entendit un jeune peintre disant avec cette enfantine légèreté qu’elle ne sut pas reconnaître et qui absout une plaisanterie de toute irréligion : — Mais, madame, votre paradis n’est pas plus beau que la Transfiguration de Raphaël ? Eh ! bien, je me suis lassé de la regarder. Augustine apporta donc dans cette société spirituelle un esprit de défiance qui n’échappait à personne. Elle gêna. Les artistes gênés sont impitoyables : ils fuient ou se moquent. Madame Guillaume avait, entre autres ridicules, celui d’outrer la dignité qui lui semblait l’apanage d’une femme mariée ; et quoiqu’elle s’en fût souvent moquée, Augustine ne sut pas se défendre d’une légère imitation de la pruderie maternelle. Cette exagération de pudeur, que n’évitent pas toujours les femmes vertueuses, suggéra quelques épigrammes à coups de crayon dont l’innocent badinage était de trop bon goût pour que Sommervieux pût s’en fâcher. Ces plaisanteries eussent été même plus cruelles, elles n’étaient après tout que des représailles exercées sur lui par ses amis. Mais rien ne pouvait être léger pour une âme qui recevait aussi facilement que celle de Théodore des impressions étrangères. Aussi éprouva-t-il insensiblement une froideur qui ne pouvait aller qu’en croissant. Pour arriver au bonheur conjugal, il faut gravir une montagne dont l’étroit plateau est bien près d’un revers aussi rapide que glissant, et l’amour du peintre le descendait. Il jugea sa femme incapable d’apprécier les considérations morales qui justifiaient, à ses propres yeux, la singularité de ses manières envers elle, et se crut fort innocent en lui cachant des pensées qu’elle ne comprenait pas et des écarts peu justifiables au tribunal d’une conscience bourgeoise. Augustine se renferma dans une douleur morne et silencieuse. Ces sentiments secrets mirent entre les deux époux un voile qui devait s’épaissir de jour en jour. Sans que son mari manquât d’égards envers elle, Augustine ne pouvait s’empêcher de trembler en le voyant réserver pour le monde les trésors d’esprit et de grâce qu’il venait jadis mettre à ses pieds. Bientôt, elle interpréta fatalement les discours spirituels qui se tiennent dans le monde sur l’inconstance des hommes. Elle ne se plaignit pas, mais son attitude équivalait à des reproches. Trois ans après son mariage, cette femme jeune et jolie qui passait si brillante dans son brillant équipage, qui vivait dans une sphère de gloire et de richesse enviée de tant de gens insouciants et incapables d’apprécier justement les situations de la vie, fut en proie à de violents chagrins. Ses couleurs pâlirent. Elle réfléchit, elle compara ; puis, le malheur lui déroula les premiers textes de l’expérience. Elle résolut de rester courageusement dans le cercle de ses devoirs, en espérant que cette conduite généreuse lui ferait recouvrer tôt ou tard l’amour de son mari ; mais il n’en fut pas ainsi. Quand Sommervieux, fatigué de travail, sortait de son atelier, Augustine ne cachait pas si promptement son ouvrage, que le peintre ne pût apercevoir sa femme raccommodant avec toute la minutie d’une bonne ménagère le linge de la maison et le sien. Elle fournissait, avec générosité, sans murmure, l’argent nécessaire aux prodigalités de son mari ; mais, dans le désir de conserver la fortune de son cher Théodore, elle se montrait économe soit pour elle, soit dans certains détails de l’administration domestique. Cette conduite est incompatible avec le laisser-aller des artistes qui, sur la fin de leur carrière, ont tant joui de la vie, qu’ils ne se demandent jamais la raison de leur ruine. Il est inutile de marquer chacune des dégradations de couleur par lesquelles la teinte brillante de leur lune de miel atteignit à une profonde obscurité. Un soir, la triste Augustine, qui depuis long-temps entendait son mari parler avec enthousiasme de madame la duchesse de Carigliano, reçut d’une amie quelques avis méchamment charitables sur la nature de l’attachement qu’avait conçu Sommervieux pour cette célèbre coquette qui donnait le ton à la cour impériale. A vingt et un ans, dans tout l’éclat de la jeunesse et de la beauté, Augustine se vit trahie pour une femme de trente-six ans. En se sentant malheureuse au milieu du monde et de ses fêtes désertes pour elle, la pauvre petite ne comprit plus rien à l’admiration qu’elle y excitait, ni à l’envie qu’elle inspirait. Sa figure prit une nouvelle expression. La mélancolie versa dans ses traits la douceur de la résignation et la pâleur d’un amour dédaigné. Elle ne tarda pas à être courtisée par les hommes les plus séduisants ; mais elle resta solitaire et vertueuse. Quelques paroles de dédain, échappées à son mari, lui donnèrent un incroyable désespoir. Une lueur fatale lui fit entrevoir les défauts de contact qui, par suite des mesquineries de son éducation, empêchaient l’union complète de son âme avec celle de Théodore : elle eut assez d’amour pour l’absoudre et pour se condamner. Elle pleura des larmes de sang, et reconnut trop tard qu’il est des mésalliances d’esprit aussi bien que des mésalliances de moeurs et de rang. En songeant aux délices printanières de son union, elle comprit l’étendue du bonheur passé, et convint en elle même qu’une si riche moisson d’amour était une vie entière qui ne pouvait se payer que par du malheur. Cependant elle aimait trop sincèrement pour perdre toute espérance. Aussi osa-t-elle entreprendre à vingt et un ans de s’instruire et de rendre son imagination au moins digne de celle qu’elle admirait.\\n\\n— Si je ne suis pas poète, se disait-elle, au moins je comprendrai la poésie.\\n\\nEt déployant alors cette force de volonté, cette énergie que les femmes possèdent toutes quand elles aiment, madame de Sommervieux tenta de changer son caractère, ses moeurs et ses habitudes ; mais en dévorant des volumes, en apprenant avec courage, elle ne réussit qu’à devenir moins ignorante. La légèreté de l’esprit et les grâces de la conversation sont un don de la nature ou le fruit d’une éducation commencée au berceau. Elle pouvait apprécier la musique, en jouir, mais non chanter avec goût. Elle comprit la littérature et les beautés de la poésie, mais il était trop tard pour en orner sa rebelle mémoire. Elle entendait avec plaisir les entretiens du monde, mais elle n’y fournissait rien de brillant. Ses idées religieuses et ses préjugés d’enfance s’opposèrent à la complète émancipation de son intelligence. Enfin, il s’était glissé contre elle, dans l’âme de Théodore, une prévention qu’elle ne put vaincre. L’artiste se moquait de ceux qui lui vantaient sa femme, et ses plaisanteries étaient assez fondées : il imposait tellement à cette jeune et touchante créature, qu’en sa présence, ou en tête-à -tête, elle tremblait. Embarrassée par son trop grand désir de plaire, elle sentait son esprit et ses connaissances s’évanouir dans un seul sentiment. La fidélité d’Augustine déplut même à cet infidèle mari, qui semblait l’engager à commettre des fautes en taxant sa vertu d’insensibilité. Augustine s’efforça en vain d’abdiquer sa raison, de se plier aux caprices, aux fantaisies de son mari, et de se vouer à l’égoïsme de sa vanité ; elle ne recueillit point le fruit de ces sacrifices. Peut-être avaient-ils tous deux laissé passer le moment où les âmes peuvent se comprendre. Un jour le coeur trop sensible de la jeune épouse reçut un de ces coups qui font si fortement plier les liens du sentiment, qu’on peut les croire rompus. Elle s’isola. Mais bientôt une fatale pensée lui suggéra d’aller chercher des consolations et des conseils au sein de sa famille.\\n\\nUn matin donc, elle se dirigea vers la grotesque façade de l’humble et silencieuse maison où s’était écoulée son enfance. Elle soupira en revoyant cette croisée d’où, un jour, elle avait envoyé un premier baiser à celui qui répandait aujourd’hui sur sa vie autant de gloire que de malheur. Rien n’était changé dans l’antre où se rajeunissait cependant le commerce de la draperie. La soeur d’Augustine occupait au comptoir antique la place de sa mère. La jeune affligée rencontra son beau-frère la plume derrière l’oreille. Elle fut à peine écoutée, tant il avait l’air affairé. Les redoutables signaux d’un inventaire général se faisaient autour de lui. Aussi la quitta-t-il en la priant d’excuser. Elle fut reçue assez froidement par sa soeur, qui lui manifesta quelque rancune. En effet, Augustine, brillante et descendant d’un joli équipage, n’était jamais venue voir sa soeur qu’en passant. La femme du prudent Lebas s’imagina que l’argent était la cause première de cette visite matinale, elle essaya de se maintenir sur un ton de réserve qui fit sourire plus d’une fois Augustine. La femme du peintre vit que, sauf les barbes au bonnet, sa mère avait trouvé dans Virginie un successeur qui conservait l’antique honneur du Chat-qui-pelote. Au déjeuner, elle aperçut, dans le régime de la maison, certains changements qui faisaient honneur au bon sens de Joseph Lebas : les commis ne se levèrent pas au dessert, on leur laissait la faculté de parler, et l’abondance de la table annonçait une aisance sans luxe. La jeune élégante trouva les coupons d’une loge aux Français où elle se souvint d’avoir vu sa soeur de loin en loin. Madame Lebas avait sur les épaules un cachemire dont la magnificence attestait la générosité avec laquelle son mari s’occupait d’elle. Enfin, les deux époux marchaient avec leur siècle. Augustine fut bientôt pénétrée d’attendrissement, en reconnaissant, pendant les deux tiers de cette journée, le bonheur égal, sans exaltation, il est vrai, mais aussi sans orages, que goûtait ce couple convenablement assorti. Ils avaient accepté la vie comme une entreprise commerciale où il s’agissait de faire, avant tout, honneur à ses affaires. La femme, n’ayant pas rencontré dans son mari un amour excessif, s’était appliquée à le faire naître. Insensiblement amené à estimer, à chérir Virginie, le temps que le bonheur mit à éclore, fut, pour Joseph Lebas et pour sa femme, un gage de durée. Aussi, lorsque la plaintive Augustine exposa sa situation douloureuse, eut-elle à essuyer le déluge de lieux communs que la morale de la rue Saint-Denis fournissait à sa soeur.\\n\\n— Le mal est fait, ma femme, dit Joseph Lebas, il faut chercher à donner de bons conseils à notre soeur. Puis, l’habile négociant analysa lourdement les ressources que les lois et les moeurs pouvaient offrir à Augustine pour sortir de cette crise ; il en numérota pour ainsi dire les considérations, les rangea par leur force dans des espèces de catégories, comme s’il se fût agi de marchandises de diverses qualités ; puis il les mit en balance, les pesa, et conclut en développant la nécessité où était sa belle-soeur de prendre un parti violent qui ne satisfit point l’amour qu’elle ressentait encore pour son mari. Aussi ce sentiment se réveilla-t-il dans toute sa force quand elle entendit Joseph Lebas parlant de voies judiciaires. Elle remercia ses deux amis, et revint chez elle encore plus indécise qu’elle ne l’était avant de les avoir consultés. Elle hasarda de se rendre alors à l’antique hôtel de la rue du Colombier, dans le dessein de confier ses malheurs à son père et à sa mère. La pauvre petite femme ressemblait à ces malades qui, arrivés à un état désespéré, essaient de toutes les recettes et se confient même aux remèdes de bonne femme. Les deux vieillards la reçurent avec une effusion de sentiment qui l’attendrit. Cette visite leur apportait une distraction qui, pour eux, valait un trésor. Depuis quatre ans, ils marchaient dans la vie comme des navigateurs sans but et sans boussole. Assis au coin de leur feu, ils se racontaient l’un à l’autre tous les désastres du Maximum, leurs anciennes acquisitions de draps, la manière dont ils avaient évité les banqueroutes, et surtout cette célèbre faillite Lecocq, la bataille de Marengo du père Guillaume. Puis, quand ils avaient épuisé les vieux procès, ils récapitulaient les additions de leurs inventaires les plus productifs, et se narraient encore les vieilles histoires du quartier Saint-Denis. A deux heures, le père Guillaume allait donner un coup d’oeil à l’établissement du Chat-qui-pelote. En revenant il s’arrêtait à toutes les boutiques, autrefois ses rivales, et dont les jeunes propriétaires espéraient entraîner le vieux négociant dans quelque escompte aventureux, que, selon sa coutume, il ne refusait jamais positivement. Deux bons chevaux normands mouraient de gras-fondu dans l’écurie de l’hôtel ; madame Guillaume ne s’en servait que pour se faire traîner tous les dimanches à la grand’messe de sa paroisse. Trois fois par semaine ce respectable couple tenait table ouverte. Grâce à l’influence de son gendre Sommervieux, le père Guillaume avait été nommé membre du comité consultatif pour l’habillement des troupes. Depuis que son mari s’était ainsi trouvé placé haut dans l’administration, madame Guillaume avait pris la détermination de représenter. Leurs appartements étaient encombrés de tant d’ornements d’or et d’argent, et de meubles sans goût mais de valeur certaine, que la pièce la plus simple y ressemblait à une chapelle. L’économie et la prodigalité semblaient se disputer dans chacun des accessoires de cet hôtel. L’on eût dit que monsieur Guillaume avait eu en vue de faire un placement d’argent jusque dans l’acquisition d’un flambeau. Au milieu de ce bazar, dont la richesse accusait le désoeuvrement des deux époux, le célèbre tableau de Sommervieux avait obtenu la place d’honneur. Il faisait la consolation de monsieur et de madame Guillaume qui tournaient vingt fois par jour leurs yeux harnachés de bésicles vers cette image de leur ancienne existence, pour eux si active et si amusante. L’aspect de cet hôtel et de ces appartements où tout avait une senteur de vieillesse et de médiocrité, le spectacle donné par ces deux êtres qui semblaient échoués sur un rocher d’or loin du monde et des idées qui font vivre, surprirent Augustine. Elle contemplait en ce moment la seconde partie du tableau dont le commencement l’avait frappée chez Joseph Lebas, celui d’une vie agitée quoique sans mouvement, espèce d’existence mécanique et instinctive semblable à celle des castors. Elle eut alors je ne sais quel orgueil de ses chagrins, en pensant qu’ils prenaient leur source dans un bonheur de dix-huit mois qui valait à ses yeux mille existences comme celle dont le vide lui semblait horrible. Cependant elle cacha ce sentiment peu charitable, et déploya pour ses vieux parents les grâces nouvelles de son esprit, les coquetteries de tendresse que l’amour lui avait révélées, et les disposa favorablement à écouter ses doléances matrimoniales. Les vieilles gens ont un faible pour ces sortes de confidences. Madame Guillaume voulut être instruite des plus légers détails de cette vie étrange qui, pour elle, avait quelque chose de fabuleux. Les voyages du baron de La Hontan, qu’elle commençait toujours sans jamais les achever, ne lui apprirent rien de plus inouï sur les sauvages du Canada.\\n\\n— Comment, mon enfant, ton mari s’enferme avec des femmes nues, et tu as la simplicité de croire qu’il les dessine ?\\n\\nA cette exclamation, la grand’mère posa ses lunettes sur une petite travailleuse, secoua ses jupons et plaça ses mains jointes sur ses genoux élevés par une chaufferette, son piédestal favori.\\n\\n— Mais, ma mère, tous les peintres sont obligés d’avoir des modèles.\\n\\n— Il s’est bien gardé de nous dire tout cela quand il t’a demandée en mariage. Si je l’avais su, je n’aurais pas donné ma fille à un homme qui fait un pareil métier. La religion défend ces horreurs-là , ça n’est pas moral. A quelle heure nous disais-tu donc qu’il rentre chez lui ?\\n\\n— Mais à une heure, deux heures…\\n\\nLes deux époux se regardèrent dans un profond étonnement.\\n\\n— Il joue donc ? dit monsieur Guillaume. Il n’y avait que les joueurs qui, de mon temps, rentrassent si tard.\\n\\nAugustine fit une petite moue qui repoussait cette accusation.\\n\\n— Il doit te faire passer de cruelles nuits à l’attendre, reprit madame Guillaume. Mais, non, tu te couches, n’est-ce pas ? Et quand il a perdu, le monstre te réveille.\\n\\n— Non, ma mère, il est au contraire quelquefois très-gai. Assez souvent même, quand il fait beau, il me propose de me lever pour aller dans les bois.\\n\\n— Dans les bois, à ces heures-là ? Tu as donc un bien petit appartement qu’il n’a pas assez de sa chambre, de ses salons, et qu’il lui faille ainsi courir pour… Mais c’est pour t’enrhumer, que le scélérat te propose ces parties-là . Il veut se débarrasser de toi. A-t-on jamais vu un homme établi, qui a un commerce tranquille, galoper comme un loup-garou ?\\n\\n— Mais, ma mère, vous ne comprenez donc pas que, pour développer son talent, il a besoin d’exaltation. Il aime beaucoup les scènes qui…\\n\\n— Ah ! je lui en ferais de belles, des scènes, moi, s’écria madame Guillaume en interrompant sa fille. Comment peux-tu garder des ménagements avec un homme pareil ? D’abord, je n’aime pas qu’il ne boive que de l’eau. Ça n’est pas sain. Pourquoi montre-t-il de la répugnance à voir les femmes quand elles mangent ? Quel singulier genre ! Mais c’est un fou. Tout ce que tu nous en as dit n’est pas possible, Un homme ne peut pas partir de sa maison sans souffler mot et ne revenir que dix jours après. Il te dit qu’il a été à Dieppe pour peindre la mer. Est-ce qu’on peint la mer ? Il te fait des contes à dormir debout.\\n\\nAugustine ouvrit la bouche pour défendre son mari ; mais madame Guillaume lui imposa silence par un geste de main auquel un reste d’habitude la fit obéir, et sa mère s’écria d’un ton sec :\\n\\n— Tiens, ne me parle pas de cet homme-là ! il n’a jamais mis le pied dans une église que pour te voir et t’épouser. Les gens sans religion sont capables de tout. Est-ce que Guillaume s’est jamais avisé de me cacher quelque chose, de rester des trois jours sans me dire ouf, et de babiller ensuite comme une pie borgne ?\\n\\n— Ma chère mère, vous jugez trop sévèrement les gens supérieurs. S’ils avaient des idées semblables à celles des autres, ce ne seraient plus des gens à talent.\\n\\n— Eh bien ! que les gens à talent restent chez eux et ne se marient pas. Comment ! un homme à talent rendra sa femme malheureuse ! et parce qu’il a du talent, ce sera bien ? Talent, talent ! Il n’y a pas tant de talent à dire comme lui blanc et noir à toute minute, à couper la parole aux gens, à battre du tambour chez soi, à ne jamais vous laisser savoir sur quel pied danser, à forcer une femme de ne pas s’amuser avant que les idées de monsieur ne soient gaies, d’être triste, dès qu’il est triste.\\n\\n— Mais, ma mère, le propre de ces imaginations-là …\\n\\n— Qu’est-ce que c’est que ces imaginations-là ? reprit madame Guillaume en interrompant encore sa fille. Il en a de belles, ma foi ! Qu’est-ce qu’un homme auquel il prend tout à coup, sans consulter de médecin, la fantaisie de ne manger que des légumes ? Encore, si c’était par religion, sa diète lui servirait à quelque chose ; mais il n’en a pas plus qu’un huguenot. A-t-on jamais vu un homme aimer, comme lui, les chevaux plus qu’il n’aime son prochain, se faire friser les cheveux comme un païen, coucher des statues sous de la mousseline, faire fermer ses fenêtres le jour pour travailler à la lampe ? Tiens, laisse-moi, s’il n’était pas si grossièrement immoral, il serait bon à mettre aux Petites-Maisons. Consulte monsieur Loraux, le vicaire de Saint-Sulpice, demande-lui son avis sur tout cela, il te dira que ton mari ne se conduit pas comme un chrétien…\\n\\n— Oh ! ma mère ! pouvez-vous croire…\\n\\n— Oui, je le crois ! Tu l’as aimé, tu n’aperçois rien de ces choses-là . Mais, moi, vers les premiers temps de son mariage, je me souviens de l’avoir rencontré dans les Champs-Elysées. Il était à cheval. Eh bien ! il galopait par moment ventre à terre, et puis il s’arrêtait pour aller pas à pas. Je me suis dit alors : « Voilà un homme qui n’a pas de jugement. »\\n\\n— Ah ! s’écria monsieur Guillaume en se frottant les mains, comme j’ai bien fait de t’avoir mariée séparée de biens avec cet original-là !\\n\\nQuand Augustine eut l’imprudence de raconter les griefs véritables qu’elle avait à exposer contre son mari, les deux vieillards restèrent muets d’indignation. Le mot de divorce fut bientôt prononcé par madame Guillaume. Au mot de divorce, l’inactif négociant fut comme réveillé. Stimulé par l’amour qu’il avait pour sa fille, et aussi par l’agitation qu’un procès allait donner à sa vie sans événements, le père Guillaume prit la parole. Il se mit à la tête de la demande en divorce, la dirigea, plaida presque, il offrit à sa fille de se charger de tous les frais, de voir les juges, les avoués, les avocats, de remuer ciel et terre. Madame de Sommervieux, effrayée, refusa les services de son père, dit qu’elle ne voulait pas se séparer de son mari, dût-elle être dix fois plus malheureuse encore, et ne parla plus de ses chagrins. Après avoir été accablée par ses parents de tous ces petits soins muets et consolateurs par lesquels les deux vieillards essayèrent de la dédommager, mais en vain, de ses peines de coeur, Augustine se retira en sentant l’impossibilité de parvenir à faire bien juger les hommes supérieurs par des esprits faibles. Elle apprit qu’une femme devait cacher à tout le monde, même à ses parents, des malheurs pour lesquels on rencontre si difficilement des sympathies. Les orages et les souffrances des sphères élevées ne peuvent être appréciés que par les nobles esprits qui les habitent. En toute chose, nous ne pouvons être jugés que par nos pairs.\\n\\nLa pauvre Augustine se retrouva donc dans la froide atmosphère de son ménage, livrée à l’horreur de ses méditations. L’étude n’était plus rien pour elle, puisque l’étude ne lui avait pas rendu le coeur de son mari. Initiée aux secrets de ces âmes de feu mais privée de leurs ressources, elle participait avec force à leurs peines sans partager leurs plaisirs. Elle s’était dégoûtée du monde, qui lui semblait mesquin et petit devant les événements des passions. Enfin, sa vie était manquée. Un soir, elle fut frappée d’une pensée qui vint illuminer ses ténébreux chagrins comme un rayon céleste. Cette idée ne pouvait sourire qu’à un coeur aussi pur, aussi vertueux que l’était le sien. Elle résolut d’aller chez la duchesse de Carigliano, non pas pour lui redemander le coeur de son mari, mais pour s’y instruire des artifices qui le lui avaient enlevé ; mais pour intéresser à la mère des enfants de son ami cette orgueilleuse femme du monde ; mais pour la fléchir et la rendre complice de son bonheur à venir comme elle était l’instrument de son malheur présent.\\n\\nUn jour donc, la timide Augustine, armée d’un courage surnaturel, monta en voiture, à deux heures après midi, pour essayer de pénétrer jusqu’au boudoir de la célèbre coquette, qui n’était jamais visible avant cette heure-là . Madame de Sommervieux ne connaissait pas encore les antiques et somptueux hôtels du faubourg Saint-Germain. Quand elle parcourut ces vestibules majestueux, ces escaliers grandioses, ces salons immenses ornés de fleurs malgré les rigueurs de l’hiver, et décorés avec ce goût particulier aux femmes qui sont nées dans l’opulence ou avec les habitudes distinguées de l’aristocratie, Augustine eut un affreux serrement de coeur. Elle envia les secrets de cette élégance de laquelle elle n’avait jamais eu l’idée. Elle respira un air de grandeur qui lui expliqua l’attrait de cette maison pour son mari. Quand elle parvint aux petits appartements de la duchesse, elle éprouva de la jalousie et une sorte de désespoir, en y admirant la voluptueuse disposition des meubles, des draperies et des étoffes tendues. Là le désordre était une grâce, là le luxe affectait une espèce de dédain pour la richesse. Les parfums répandus dans cette douce atmosphère flattaient l’odorat sans l’offenser. Les accessoires de l’appartement s’harmoniaient avec une vue ménagée par des glaces sans tain sur les pelouses d’un jardin planté d’arbres verts. Tout était séduction, et le calcul ne s’y sentait point. Le génie de la maîtresse de ces appartements respirait tout entier dans le salon où attendait Augustine. Elle tâcha d’y deviner le caractère de sa rivale par l’aspect des objets épars ; mais il y avait là quelque chose d’impénétrable dans le désordre comme dans la symétrie, et pour la simple Augustine ce fut lettres closes. Tout ce qu’elle put y voir, c’est que la duchesse était une femme supérieure en tant que femme. Elle eut alors une pensée douloureuse.\\n\\n— Hélas ! serait-il vrai, se dit-elle, qu’un coeur aimant et simple ne suffit pas à un artiste ; et pour balancer le poids de ces âmes fortes, faut-il les unir à des âmes féminines dont la puissance soit pareille à la leur ? Si j’avais été élevée comme cette sirène, au moins nos armes eussent été égales au moment de la lutte.\\n\\n— Mais je n’y suis pas ! Ces mots secs et brefs, quoique prononcés à voix basse dans le boudoir voisin, furent entendus par Augustine, dont le coeur palpita.\\n\\n— Cette dame est là , répliqua la femme de chambre.\\n\\n— Vous êtes folle, faites donc entrer ! répondit la duchesse dont la voix devenue douce avait pris l’accent affectueux de la politesse. Evidemment, elle désirait alors être entendue.\\n\\nAugustine s’avança timidement. Au fond de ce frais boudoir elle vit la duchesse voluptueusement couchée sur une ottomane en velours vert placée au centre d’une espèce de demi-cercle dessiné par les plis moelleux d’une mousseline tendue sur un fond jaune. Des ornements de bronze doré, disposés avec un goût exquis, rehaussaient encore cette espèce de dais sous lequel la duchesse était posée comme une statue antique. La couleur foncée du velours ne lui laissait perdre aucun moyen de séduction. Un demi-jour, ami de sa beauté, semblait être plutôt un reflet qu’une lumière. Quelques fleurs rares élevaient leurs têtes embaumées au dessus des vases de Sèvres les plus riches. Au moment où ce tableau s’offrit aux yeux d’Augustine étonnée, elle avait marché si doucement, qu’elle put surprendre un regard de l’enchanteresse. Ce regard semblait dire à une personne que la femme du peintre n’aperçut pas d’abord :\\n\\n— Restez, vous allez voir une jolie femme, et vous me rendrez sa visite moins ennuyeuse.\\n\\nA l’aspect d’Augustine, la duchesse se leva et la fit asseoir auprès d’elle.\\n\\n— A quoi dois-je le bonheur de cette visite, madame ? dit-elle avec un sourire plein de grâces.\\n\\n— Pourquoi tant de fausseté ? pensa Augustine, qui ne répondit que par une inclination de tête.\\n\\nCe silence était commandé. La jeune femme voyait devant elle un témoin de trop à cette scène. Ce personnage était, de tous les colonels de l’armée, le plus jeune, le plus élégant et le mieux fait. Son costume demi-bourgeois faisait ressortir les grâces de sa personne. Sa figure pleine de vie, de jeunesse, et déjà fort expressive, était encore animée par de petites moustaches relevées en pointe et noires comme du jais, par une impériale bien fournie, par des favoris soigneusement peignés et par une forêt de cheveux noirs assez en désordre. Il badinait avec une cravache, en manifestant une aisance et une liberté qui seyaient à l’air satisfait de sa physionomie ainsi qu’à la recherche de sa toilette. Les rubans attachés à sa boutonnière étaient noués avec dédain, et il paraissait bien plus vain de sa jolie tournure que de son courage. Augustine regarda la duchesse de Carigliano en lui montrant le colonel par un coup d’oeil dont toutes les prières furent comprises.\\n\\n— Eh bien, adieu, monsieur d’Aiglemont, nous nous retrouverons au bois de Boulogne.\\n\\nCes mots furent prononcés par la sirène comme s’ils étaient le résultat d’une stipulation antérieure à l’arrivée d’Augustine ; elle les accompagna d’un regard menaçant que l’officier méritait peut-être pour l’admiration qu’il témoignait en contemplant la modeste fleur qui contrastait si bien avec l’orgueilleuse duchesse. Le jeune fat s’inclina en silence, tourna sur les talons de ses bottes, et s’élança gracieusement hors du boudoir. En ce moment, Augustine, épiant sa rivale qui semblait suivre des yeux le brillant officier, surprit dans ce regard un sentiment dont les fugitives expressions sont connues de toutes les femmes. Elle songea avec la douleur la plus profonde que sa visite allait être inutile : cette artificieuse duchesse était trop avide d’hommages pour ne pas avoir le coeur sans pitié.\\n\\n— Madame, dit Augustine d’une voix entrecoupée, la démarche que je fais en ce moment auprès de vous va vous sembler bien singulière ; mais le désespoir a sa folie, et doit faire tout excuser. Je m’explique trop bien pourquoi Théodore préfère votre maison à toute autre, et pourquoi votre esprit exerce tant d’empire sur lui. Hélas ! je n’ai qu’à rentrer en moi-même pour en trouver des raisons plus que suffisantes. Mais j’adore mon mari, madame. Deux ans de larmes n’ont point effacé son image de mon coeur, quoique j’aie perdu le sien. Dans ma folie, j’ai osé concevoir l’idée de lutter avec vous ; et je viens à vous, vous demander par quels moyens je puis triompher de vous-même. Oh, madame ! s’écria la jeune femme en saisissant avec ardeur la main de sa rivale, qui la lui laissa prendre, je ne prierai jamais Dieu pour mon propre bonheur avec autant de ferveur que je l’implorerais pour le vôtre, si vous m’aidiez à reconquérir, je ne dirai pas l’amour, mais la tendresse de Sommervieux. Je n’ai plus d’espoir qu’en vous. Ah ! dites-moi comment vous avez pu lui plaire et lui faire oublier les premiers jours de…\\n\\nA ces mots, Augustine, suffoquée par des sanglots mal contenus, fut obligée de s’arrêter. Honteuse de sa faiblesse, elle cacha son visage dans un mouchoir qu’elle inonda de ses larmes.\\n\\n— Etes-vous donc enfant, ma chère petite belle ! dit la duchesse, qui, séduite par la nouveauté de cette scène et attendrie malgré elle en recevant l’hommage que lui rendait la plus parfaite vertu qui fût peut-être à Paris, prit le mouchoir de la jeune femme et se mit à lui essuyer elle-même les yeux en la flattant par quelques monosyllabes murmurés avec une gracieuse pitié.\\n\\nAprès un moment de silence, la coquette, emprisonnant les jolies mains de la pauvre Augustine entre les siennes qui avaient un rare caractère de beauté noble et de puissance, lui dit d’une voix douce et affectueuse :\\n\\n— Pour premier avis, je vous conseillerai de ne pas pleurer ainsi, les larmes enlaidissent. Il faut savoir prendre son parti sur les chagrins ; ils rendent malade, et l’amour ne reste pas long-temps sur un lit de douleur. La mélancolie donne bien d’abord une certaine grâce qui plaît ; mais elle finit par allonger les traits et flétrir la plus ravissante de toutes les figures. Ensuite, nos tyrans ont l’amour-propre de vouloir que leurs esclaves soient toujours gaies.\\n\\n— Ah, madame ! il ne dépend pas de moi de ne pas sentir ! Comment peut-on, sans éprouver mille morts, voir terne, décolorée, indifférente, une figure qui jadis rayonnait d’amour et de joie ? Ah ! je ne sais pas commander à mon coeur.\\n\\n— Tant pis, chère belle ; mais je crois déjà savoir toute votre histoire. D’abord, imaginez-vous bien que si votre mari vous a été infidèle, je ne suis pas sa complice. Si j’ai tenu à l’avoir dans mon salon, c’est, je l’avouerai, par amour-propre : il était célèbre et n’allait nulle part. Je vous aime déjà trop pour vous dire toutes les folies qu’il a faites pour moi. Je ne vous en révélerai qu’une seule, parce qu’elle nous servira peut-être à vous le ramener et à le punir de l’audace qu’il met dans ses procédés avec moi. Il finirait par me compromettre. Je connais trop le monde, ma chère, pour vouloir me mettre à la discrétion d’un homme trop supérieur. Sachez qu’il faut se laisser faire la cour par eux, mais les épouser ! c’est une faute. Nous autres femmes, nous devons admirer les hommes de génie, en jouir comme d’un spectacle, mais vivre avec eux ! jamais. Fi donc ! c’est vouloir prendre plaisir à regarder les machines de l’opéra, au lieu de rester dans une loge, à y savourer ses brillantes illusions. Mais chez vous, ma pauvre enfant, le mal est arrivé, n’est-ce pas ? Eh bien ! il faut essayer de vous armer contre la tyrannie.\\n\\n— Ah, madame ! avant d’entrer ici, en vous y voyant, j’ai déjà reconnu quelques artifices que je ne soupçonnais pas.\\n\\n— Eh bien, venez me voir quelquefois, et vous ne serez pas long-temps sans posséder la science de ces bagatelles, d’ailleurs assez importantes. Les choses extérieures sont, pour les sots, la moitié de la vie ; et pour cela, plus d’un homme de talent se trouve un sot malgré tout son esprit. Mais je gage que vous n’avez jamais rien su refuser à Théodore ?\\n\\n— Le moyen, madame, de refuser quelque chose à celui qu’on aime !\\n\\n— Pauvre innocente, je vous adorerais pour votre niaiserie. Sachez donc que plus nous aimons, moins nous devons laisser apercevoir à un homme, surtout à un mari, l’étendue de notre passion. C’est celui qui aime le plus qui est tyrannisé, et, qui pis est, délaissé tôt ou tard. Celui qui veut régner, doit…\\n\\n— Comment, madame ! faudra-t-il donc dissimuler, calculer, devenir fausse, se faire un caractère artificiel et pour toujours ? Oh ! comment peut-on vivre ainsi ? Est-ce que vous pouvez…\\n\\nElle hésita, la duchesse sourit.\\n\\n— Ma chère, reprit la grande dame d’une voix grave, le bonheur conjugal a été de tout temps une spéculation, une affaire qui demande une attention particulière. Si vous continuez à parler passion quand je vous parle mariage, nous ne nous entendrons bientôt plus. Ecoutez-moi, continua-t-elle en prenant le ton d’une confidence. J’ai été à même de voir quelques-uns des hommes supérieurs de notre époque. Ceux qui se sont mariés ont, à quelques exceptions près, épousé des femmes nulles. Eh bien ! ces femmes-là les gouvernaient, comme l’empereur nous gouverne, et étaient, sinon aimées, du moins respectées par eux. J’aime assez les secrets, surtout ceux qui nous concernent, pour m’être amusée à chercher le mot de cette énigme. Eh bien, mon ange ! ces bonnes femmes avaient le talent d’analyser le caractère de leurs maris. Sans s’épouvanter comme vous de leurs supériorités, elles avaient adroitement remarqué les qualités qui leur manquaient. Soit qu’elles possédassent ces qualités, ou qu’elles feignissent de les avoir, elles trouvaient moyen d’en faire un si grand étalage aux yeux de leurs maris qu’elles finissaient par leur imposer. Enfin, apprenez encore que ces âmes qui paraissent si grandes ont toutes un petit grain de folie que nous devons savoir exploiter. En prenant la ferme volonté de les dominer, en ne s’écartant jamais de ce but, en y rapportant toutes nos actions, nos idées, nos coquetteries, nous maîtrisons ces esprits éminemment capricieux qui, par la mobilité même de leurs pensées, nous donnent les moyens de les influencer.\\n\\n— Oh ciel ! s’écria la jeune femme épouvantée, voilà donc la vie.\\n\\nC’est un combat…\\n\\n\\n\\n— Où il faut toujours menacer, reprit la duchesse en riant. Notre pouvoir est tout factice. Aussi ne faut-il jamais se laisser mépriser par un homme ; on ne se relève d’une pareille chute que par des manoeuvres odieuses. Venez, ajouta-t-elle, je vais vous donner un moyen de mettre votre mari à la chaîne.\\n\\nElle se leva, pour guider en souriant la jeune et innocente apprentie des ruses conjugales à travers le dédale de son petit palais. Elles arrivèrent toutes deux à un escalier dérobé qui communiquait aux appartements de réception. Quand la duchesse tourna le secret de la porte, elle s’arrêta, regarda Augustine avec un air inimitable de finesse et de grâce :\\n\\n— Tenez, le duc de Carigliano m’adore ! eh bien, il n’ose pas entrer par cette porte sans ma permission. Et c’est un homme qui a l’habitude de commander à des milliers de soldats. Il sait affronter les batteries, mais devant moi ! il a peur.\\n\\nAugustine soupira. Elles parvinrent à une somptueuse galerie où la femme du peintre fut amenée par la duchesse devant le portrait que Théodore avait fait de mademoiselle Guillaume. A cet aspect, Augustine jeta un cri.\\n\\n— Je savais bien qu’il n’était plus chez moi, dit-elle, mais… ici !\\n\\n— Ma chère, je ne l’ai exigé que pour voir jusqu’à quel degré de bêtise un homme de génie peut atteindre. Tôt ou tard, il vous aurait été rendu par moi ; mais je ne m’attendais pas au plaisir de voir ici l’original devant la copie. Pendant que nous allons achever notre conversation, je le ferai porter dans votre voiture. Si, armée de ce talisman, vous n’êtes pas maîtresse de votre mari pendant cent ans, vous n’êtes pas une femme, et vous méritez votre sort !\\n\\nAugustine baisa la main de la duchesse, qui la pressa sur son coeur et l’embrassa avec une tendresse d’autant plus vive qu’elle devait être oubliée le lendemain. Cette scène aurait peut-être à jamais ruiné la candeur et la pureté d’une femme moins vertueuse qu’Augustine, à qui les secrets révélés par la duchesse pouvaient être également salutaires et funestes. La politique astucieuse des hautes sphères sociales ne convenait pas plus à Augustine que l’étroite raison de Joseph Lebas, ou que la niaise morale de madame Guillaume. Etrange effet des fausses positions où nous jettent les moindres contresens commis dans la vie ! Augustine ressemblait alors à un pâtre des Alpes surpris par une avalanche : s’il hésite, ou s’il veut écouter les cris de ses compagnons, le plus souvent il périt. Dans ces grandes crises, le coeur se brise ou se bronze.\\n\\nMadame de Sommervieux revint chez elle en proie à une agitation qu’il serait difficile de décrire. Sa conversation avec la duchesse de Carigliano éveillait une foule d’idées contradictoires dans son esprit. Elle était comme les moutons de la fable, pleine de courage en l’absence du loup. Elle se haranguait elle-même et se traçait d’admirables plans de conduite ; elle concevait mille stratagèmes de coquetterie ; elle parlait même à son mari, retrouvant, loin de lui, toutes les ressources de cette éloquence vraie qui n’abandonne jamais les femmes ; puis, en songeant au regard fixe et clair de Théodore, elle tremblait déjà . Quand elle demanda si monsieur était chez lui, la voix lui manqua. En apprenant qu’il ne reviendrait pas dîner, elle éprouva un mouvement de joie inexplicable. Semblable au criminel qui se pourvoit en cassation contre son arrêt de mort, un délai, quelque court qu’il pût être, lui semblait une vie entière. Elle plaça le portrait dans sa chambre, et attendit son mari en se livrant à toutes les angoisses de l’espérance Elle pressentait trop bien que cette tentative allait décider de tout son avenir, pour ne pas frissonner à toute espèce de bruit, même au murmure de sa pendule qui semblait appesantir ses terreurs en les lui mesurant. Elle tâcha de tromper le temps par mille artifices. Elle eut l’idée de faire une toilette qui la rendit semblable en tout point au portrait. Puis, connaissant le caractère inquiet de son mari, elle fit éclairer son appartement d’une manière inusitée, certaine qu’en rentrant la curiosité l’amènerait chez elle. Minuit sonna, quand, au cri du jockey, la porte de l’hôtel s’ouvrit. La voiture du peintre roula sur le pavé de la cour silencieuse.\\n\\n— Que signifie cette illumination ? demanda Théodore d’une voix joyeuse en entrant dans la chambre de sa femme.\\n\\nAugustine saisit avec adresse un moment si favorable, elle s’élança au cou de son mari et lui montra le portrait. L’artiste resta immobile comme un rocher. Ses yeux se dirigèrent alternativement sur Augustine et sur la toile accusatrice. La timide épouse, demi-morte, épiait le front changeant, le front terrible de son mari. Elle en vit par degrés les rides expressives s’amonceler comme des nuages ; puis, elle crut sentir son sang se figer dans ses veines, quand, par un regard flamboyant et d’une voix profondément sourde, elle fut interrogée.\\n\\n— Où avez-vous trouvé ce tableau ?\\n\\n— La duchesse de Carigliano me l’a rendu.\\n\\n— Vous le lui avez demandé ?\\n\\n— Je ne savais pas qu’il fût chez elle.\\n\\nLa douceur ou plutôt la mélodie enchanteresse de la voix de cet ange eût attendri des Cannibales, mais non un artiste en proie aux tortures de la vanité blessée.\\n\\n— Cela est digne d’elle, s’écria l’artiste d’une voix tonnante. Je me vengerai ! dit-il en se promenant à grands pas. Elle en mourra de honte : je la peindrai ! oui, je la représenterai sous les traits de Messaline sortant à la nuit du palais de Claude.\\n\\n— Théodore ! dit une voix mourante.\\n\\n— Je la tuerai.\\n\\n— Mon ami !\\n\\n— Elle aime ce petit colonel de cavalerie, parce qu’il monte bien à cheval…\\n\\n— Théodore !\\n\\n— Eh ! laissez-moi, dit le peintre à sa femme avec un son de voix qui ressemblait presque à un rugissement.\\n\\nIl serait odieux de peindre toute cette scène à la fin de laquelle l’ivresse de la colère suggéra à l’artiste des paroles et des actes qu’une femme, moins jeune qu’Augustine, aurait attribués à la démence.\\n\\nSur les huit heures du matin, le lendemain, madame Guillaume surprit sa fille pâle, les yeux rouges, la coiffure en désordre, tenant à la main un mouchoir trempé de pleurs, contemplant sur le parquet les fragments épars d’une toile déchirée et les morceaux d’un grand cadre doré mis en pièce. Augustine, que la douleur rendait presque insensible, montra ces débris par un geste empreint de désespoir.\\n\\n— Et voilà peut-être une grande perte, s’écria la vieille régente du Chat-qui-pelote. Il était ressemblant, c’est vrai ; mais j’ai appris qu’il y a sur le boulevard un homme qui fait des portraits charmants pour cinquante écus.\\n\\n— Ah, ma mère !\\n\\n— Pauvre petite, tu as bien raison ! répondit madame Guillaume qui méconnut l’expression du regard que lui jeta sa fille. Va, mon enfant, l’on n’est jamais si tendrement aimé que par sa mère. Ma mignonne, je devine tout ; mais viens me confier tes chagrins, je te consolerai. Ne t’ai-je pas déjà dit que cet homme-là était un fou ! Ta femme de chambre m’a conté de belles choses… Mais c’est donc un véritable monstre !\\n\\nAugustine mit un doigt sur ses lèvres pâlies, comme pour implorer de sa mère un moment de silence. Pendant cette terrible nuit, le malheur lui avait fait trouver cette patiente résignation qui, chez les mères et chez les femmes aimantes, surpasse, dans ses effets, l’énergie humaine et révèle peut-être dans le coeur des femmes l’existence de certaines cordes que Dieu a refusées à l’homme.\\n\\nUne inscription gravée sur un cippe du cimetière Montmartre indiquait que madame de Sommervieux était morte à vingt-sept ans. Un poète, ami de cette timide créature, voyait, dans les simples lignes de son épitaphe, la dernière scène d’un drame. Chaque année, au jour solennel du 2 novembre, il ne passait jamais devant ce jeune marbre sans se demander s’il ne fallait pas des femmes plus fortes que ne l’était Augustine pour les puissantes étreintes du génie.\\n\\n— Les humbles et modestes fleurs, écloses dans les vallées, meurent peut-être, se disait-il, quand elles sont transplantées trop près des cieux, aux régions où se forment les orages, où le soleil est brûlant.\\n\\nMaffliers, octobre 1829.\\n\\n\\n\\n'"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 54
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "OZ9E53RUjKRm"
+ },
+ "source": [
+ "## Transform novels to sentence lists"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "cWCEbPxVjKRm",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 72
+ },
+ "outputId": "364f2b80-5631-4b42-b4e0-4dbe21176ca0"
+ },
+ "source": [
+ "import nltk\n",
+ "nltk.download('punkt')\n"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
+ "[nltk_data] Unzipping tokenizers/punkt.zip.\n"
+ ],
+ "name": "stdout"
+ },
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "True"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 55
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Hi5OSTtJjKRo"
+ },
+ "source": [
+ "#Fonction pour diviser le texte aux phrases\n",
+ "import re\n",
+ "alphabets= \"([A-Za-z])\"\n",
+ "prefixes = \"(Mr|St|Mrs|Ms|Dr)[.]\"\n",
+ "suffixes = \"(Inc|Ltd|Jr|Sr|Co)\"\n",
+ "starters = \"(Mr|Mrs|Ms|Dr|He\\s|She\\s|It\\s|They\\s|Their\\s|Our\\s|We\\s|But\\s|However\\s|That\\s|This\\s|Wherever)\"\n",
+ "acronyms = \"([A-Z][.][A-Z][.](?:[A-Z][.])?)\"\n",
+ "websites = \"[.](com|net|org|io|gov)\"\n",
+ "\n",
+ "def split_into_sentences(text):\n",
+ " text = \" \" + text + \" \"\n",
+ " text = text.replace(\"\\n\",\" \")\n",
+ " text = re.sub(prefixes,\"\\\\1<prd>\",text)\n",
+ " text = re.sub(websites,\"<prd>\\\\1\",text)\n",
+ " if \"Ph.D\" in text: text = text.replace(\"Ph.D.\",\"Ph<prd>D<prd>\")\n",
+ " text = re.sub(\"\\s\" + alphabets + \"[.] \",\" \\\\1<prd> \",text)\n",
+ " text = re.sub(acronyms+\" \"+starters,\"\\\\1<stop> \\\\2\",text)\n",
+ " text = re.sub(alphabets + \"[.]\" + alphabets + \"[.]\" + alphabets + \"[.]\",\"\\\\1<prd>\\\\2<prd>\\\\3<prd>\",text)\n",
+ " text = re.sub(alphabets + \"[.]\" + alphabets + \"[.]\",\"\\\\1<prd>\\\\2<prd>\",text)\n",
+ " text = re.sub(\" \"+suffixes+\"[.] \"+starters,\" \\\\1<stop> \\\\2\",text)\n",
+ " text = re.sub(\" \"+suffixes+\"[.]\",\" \\\\1<prd>\",text)\n",
+ " text = re.sub(\" \" + alphabets + \"[.]\",\" \\\\1<prd>\",text)\n",
+ " if \"â€\" in text: text = text.replace(\".â€\",\"â€.\")\n",
+ " if \"\\\"\" in text: text = text.replace(\".\\\"\",\"\\\".\")\n",
+ " if \"!\" in text: text = text.replace(\"!\\\"\",\"\\\"!\")\n",
+ " if \"?\" in text: text = text.replace(\"?\\\"\",\"\\\"?\")\n",
+ " text = text.replace(\".\",\".<stop>\")\n",
+ " text = text.replace(\"?\",\"?<stop>\")\n",
+ " text = text.replace(\"!\",\"!<stop>\")\n",
+ " text = text.replace(\"<prd>\",\".\")\n",
+ " sentences = text.split(\"<stop>\")\n",
+ " sentences = sentences[:-1]\n",
+ " sentences = [s.strip() for s in sentences]\n",
+ " return sentences"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "xT6XWeDEjKRq"
+ },
+ "source": [
+ "splited_sentences= [None] * len(content_french)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "hsx3TbevjKRv"
+ },
+ "source": [
+ "for i in range(len(content_french)):\n",
+ " splited_sentences[i]=split_into_sentences(content_french[i])\n",
+ "\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "469uhMnHjKRz",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "584f765c-b895-4401-de02-86d5dc1b8252"
+ },
+ "source": [
+ "for i in range(len(splited_sentences)):\n",
+ " print(\"nombre de phrases du roman \" + onlyfiles[i], len(splited_sentences[i]))"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "nombre de phrases du roman maison.txt 1010\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Giy2Kln7jKR1"
+ },
+ "source": [
+ "test_sentences = [None] * len(splited_sentences)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "DIk7E8fXjKR4"
+ },
+ "source": [
+ "#Convert the lists of sentences to ndarray by numpy\n",
+ "for i in range(len(splited_sentences)):\n",
+ " test_sentences[i] = np.asarray(splited_sentences[i], dtype=np.str)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "xM_PqDMHjKR5",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "c3d50004-8e82-4226-ee50-93951d030b02"
+ },
+ "source": [
+ "onlyfiles[0]"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "'maison.txt'"
+ ]
+ },
+ "metadata": {
+ "tags": []
+ },
+ "execution_count": 63
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "0dOVLPowjKR8"
+ },
+ "source": [
+ "import pandas as pd\n",
+ "\n",
+ "#input_ids_test = [None] * len(test_sentences)\n",
+ "#attention_masks = [None] * len()\n",
+ "prediction_dataloader = [None] * len(test_sentences)\n",
+ "\n",
+ "for i in range(len(test_sentences)):\n",
+ " # Tokenize all of the sentences and map the tokens to thier word IDs.\n",
+ " #input_ids_test[i] = [None] * len(test_sentences[i])\n",
+ " #prediction_dataloader[i] = [None] * len(test_sentences[i])\n",
+ " input_ids_test = []\n",
+ " # For every sentence...\n",
+ " for sent in test_sentences[i]:\n",
+ " # `encode` will:\n",
+ " # (1) Tokenize the sentence.\n",
+ " # (2) Prepend the `[CLS]` token to the start.\n",
+ " # (3) Append the `[SEP]` token to the end.\n",
+ " # (4) Map tokens to their IDs.\n",
+ " encoded_sent = tokenizer.encode(\n",
+ " sent, # Sentence to encode.\n",
+ " add_special_tokens = True, # Add '[CLS]' and '[SEP]'\n",
+ " )\n",
+ " \n",
+ " input_ids_test.append(encoded_sent)\n",
+ "\n",
+ " # Pad our input tokens\n",
+ " input_ids_test = pad_sequences(input_ids_test, maxlen=MAX_LEN, \n",
+ " dtype=\"long\", truncating=\"post\", padding=\"post\")\n",
+ "\n",
+ " # Create attention masks\n",
+ " attention_masks = []\n",
+ "\n",
+ " # Create a mask of 1s for each token followed by 0s for padding\n",
+ " for seq in input_ids_test:\n",
+ " seq_mask = [float(i>0) for i in seq]\n",
+ " attention_masks.append(seq_mask) \n",
+ "\n",
+ " # Convert to tensors.\n",
+ " prediction_inputs = torch.tensor(input_ids_test)\n",
+ " prediction_masks = torch.tensor(attention_masks)\n",
+ " #prediction_labels = torch.tensor(labels)\n",
+ "\n",
+ " # Set the batch size. \n",
+ " batch_size = 32 \n",
+ "\n",
+ " # Create the DataLoader.\n",
+ " prediction_data = TensorDataset(prediction_inputs, prediction_masks)\n",
+ " prediction_sampler = SequentialSampler(prediction_data)\n",
+ " prediction_dataloader[i] = DataLoader(prediction_data, sampler=prediction_sampler, batch_size=batch_size)\n",
+ "\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Zs-Km9IhjKSC",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 54
+ },
+ "outputId": "c5dd8217-68ff-48e1-da15-ae613da993d3"
+ },
+ "source": [
+ "predictions = []\n",
+ "# Prediction on novels\n",
+ "for i in range(len(prediction_dataloader)):\n",
+ " print('Predicting labels for {:,} test sentences '.format(len(test_sentences[i])) + 'for the novel ' + onlyfiles[i])\n",
+ "\n",
+ " # Put model in evaluation mode\n",
+ " model.eval()\n",
+ "\n",
+ " # Tracking variables \n",
+ " predictions_i = []\n",
+ "\n",
+ " # Predict \n",
+ " for batch in prediction_dataloader[i]:\n",
+ " # Add batch to GPU\n",
+ " batch = tuple(t.to(device) for t in batch)\n",
+ " \n",
+ " # Unpack the inputs from our dataloader\n",
+ " b_input_ids, b_input_mask= batch\n",
+ " \n",
+ " # Telling the model not to compute or store gradients, saving memory and \n",
+ " # speeding up prediction\n",
+ " with torch.no_grad():\n",
+ " # Forward pass, calculate logit predictions\n",
+ " outputs = model(b_input_ids, token_type_ids=None, \n",
+ " attention_mask=b_input_mask)\n",
+ " #print(outputs[0])\n",
+ " logits = outputs[0]\n",
+ "\n",
+ " # Move logits and labels to CPU\n",
+ " logits = logits.detach().cpu().numpy()\n",
+ " #label_ids = b_labels.to('cpu').numpy()\n",
+ " \n",
+ " # Store predictions and true labels\n",
+ " predictions_i.append(logits)\n",
+ " #print(predictions_i)\n",
+ " predictions.append(predictions_i)\n",
+ " #true_labels.append(label_ids)\n",
+ "\n",
+ "print(' DONE.')"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Predicting labels for 1,010 test sentences for the novel maison.txt\n",
+ " DONE.\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "RWSek-upjKSF",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 35
+ },
+ "outputId": "d855d2e8-a565-4ef4-f8b5-d5553a94de47"
+ },
+ "source": [
+ "pred_labels = []\n",
+ "\n",
+ "for j in range(len(prediction_dataloader)): \n",
+ " # Evaluate each test batch using Matthew's correlation coefficient\n",
+ " print('Calculating Matthews Corr. Coef. for each batch of ' + onlyfiles[j])\n",
+ " pred_labels_i= []\n",
+ " # For each input batch...\n",
+ " for i in range(len(prediction_dataloader[j])):\n",
+ " # The predictions for this batch are a 2-column ndarray (one column for \"0\" \n",
+ " # and one column for \"1\"). Pick the label with the highest value and turn this\n",
+ " # in to a list of 0s and 1s.\n",
+ " pred_labels_i_j = np.argmax(predictions[j][i], axis=1).flatten()\n",
+ " pred_labels_i.append(pred_labels_i_j)\n",
+ " pred_labels.append(pred_labels_i)"
+ ],
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "text": [
+ "Calculating Matthews Corr. Coef. for each batch of maison.txt\n"
+ ],
+ "name": "stdout"
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "YTBIBqgfjKSH"
+ },
+ "source": [
+ "pred_rom_list = []\n",
+ "for pred_rom in range(len(pred_labels)):\n",
+ " pred_rom_i_list = []\n",
+ " for pred_bat in range(len(pred_labels[pred_rom])):\n",
+ " pred_rom_i_list.append(pred_labels[pred_rom][pred_bat].tolist())\n",
+ " pred_rom_list.append(pred_rom_i_list)"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "uIRz3KGXjKSI"
+ },
+ "source": [
+ "roms_list_fin= []\n",
+ "for i in range(len(pred_rom_list)):\n",
+ " flat_list = [item for sublist in pred_rom_list[i] for item in sublist]\n",
+ " roms_list_fin.append(flat_list)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "5q_vkX4fjKSJ"
+ },
+ "source": [
+ "# on va voir dans les résultats parfois 1 pour une phrase qui n'indique pas sentences Geo.\n",
+ "# mais cette phrase est liée soit à une phrase Geo avant soit à une phrase Geo après\n",
+ "import pandas\n",
+ "for i in range(len(onlyfiles)):\n",
+ " df = pandas.DataFrame(data={\"sentences\": test_sentences[i], \"labels\": roms_list_fin[i]})\n",
+ " df.to_csv(\"./resultats_\" + onlyfiles[i] + \".csv\", sep=',',index=False)\n"
+ ],
+ "execution_count": null,
+ "outputs": []
+ }
+ ]
+}
\ No newline at end of file
diff --git a/notebooks/ambiguity.ipynb b/notebooks/ambiguity.ipynb
new file mode 100644
index 0000000..94615cf
--- /dev/null
+++ b/notebooks/ambiguity.ipynb
@@ -0,0 +1,387 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from helpers import read_geonames_2"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 26,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "df = read_geonames_2(\"data/FR/FR.txt\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 27,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>country_code</th>\n",
+ " <th>postal_code</th>\n",
+ " <th>placename</th>\n",
+ " <th>admin_name1</th>\n",
+ " <th>admin_code1</th>\n",
+ " <th>admin_name2</th>\n",
+ " <th>admin_code2</th>\n",
+ " <th>admin_name3</th>\n",
+ " <th>admin_code3</th>\n",
+ " <th>latitude</th>\n",
+ " <th>longitude</th>\n",
+ " <th>accuracy</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>FR</td>\n",
+ " <td>75000</td>\n",
+ " <td>Paris</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>11</td>\n",
+ " <td>Paris</td>\n",
+ " <td>75</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>48.8534</td>\n",
+ " <td>2.3488</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>FR</td>\n",
+ " <td>75001</td>\n",
+ " <td>Paris 01</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>11</td>\n",
+ " <td>Paris</td>\n",
+ " <td>75</td>\n",
+ " <td>Paris</td>\n",
+ " <td>751</td>\n",
+ " <td>48.8592</td>\n",
+ " <td>2.3417</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>FR</td>\n",
+ " <td>75001</td>\n",
+ " <td>Paris</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>11</td>\n",
+ " <td>Paris</td>\n",
+ " <td>75</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>48.8534</td>\n",
+ " <td>2.3488</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>FR</td>\n",
+ " <td>75002</td>\n",
+ " <td>Paris 02</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>11</td>\n",
+ " <td>Paris</td>\n",
+ " <td>75</td>\n",
+ " <td>Paris</td>\n",
+ " <td>751</td>\n",
+ " <td>48.8655</td>\n",
+ " <td>2.3426</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>FR</td>\n",
+ " <td>75002</td>\n",
+ " <td>Paris</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>11</td>\n",
+ " <td>Paris</td>\n",
+ " <td>75</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>48.8534</td>\n",
+ " <td>2.3488</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>...</th>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>51662</th>\n",
+ " <td>FR</td>\n",
+ " <td>20601 CEDEX</td>\n",
+ " <td>Bastia</td>\n",
+ " <td>Corse</td>\n",
+ " <td>94</td>\n",
+ " <td>Haute-Corse</td>\n",
+ " <td>2B</td>\n",
+ " <td>Arrondissement de Bastia</td>\n",
+ " <td>2B2</td>\n",
+ " <td>42.7028</td>\n",
+ " <td>9.4500</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>51663</th>\n",
+ " <td>FR</td>\n",
+ " <td>20604 CEDEX</td>\n",
+ " <td>Bastia</td>\n",
+ " <td>Corse</td>\n",
+ " <td>94</td>\n",
+ " <td>Haute-Corse</td>\n",
+ " <td>2B</td>\n",
+ " <td>Arrondissement de Bastia</td>\n",
+ " <td>2B2</td>\n",
+ " <td>42.7028</td>\n",
+ " <td>9.4500</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>51664</th>\n",
+ " <td>FR</td>\n",
+ " <td>20611 CEDEX</td>\n",
+ " <td>Bastia</td>\n",
+ " <td>Corse</td>\n",
+ " <td>94</td>\n",
+ " <td>Haute-Corse</td>\n",
+ " <td>2B</td>\n",
+ " <td>Arrondissement de Bastia</td>\n",
+ " <td>2B2</td>\n",
+ " <td>42.7028</td>\n",
+ " <td>9.4500</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>51665</th>\n",
+ " <td>FR</td>\n",
+ " <td>20620</td>\n",
+ " <td>Biguglia</td>\n",
+ " <td>Corse</td>\n",
+ " <td>94</td>\n",
+ " <td>Haute-Corse</td>\n",
+ " <td>2B</td>\n",
+ " <td>Arrondissement de Bastia</td>\n",
+ " <td>2B2</td>\n",
+ " <td>42.6269</td>\n",
+ " <td>9.4202</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>51666</th>\n",
+ " <td>FR</td>\n",
+ " <td>98799</td>\n",
+ " <td>Clipperton Island</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>10.2922</td>\n",
+ " <td>-109.2072</td>\n",
+ " <td>5.0</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "<p>51667 rows × 12 columns</p>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " country_code postal_code placename admin_name1 admin_code1 \\\n",
+ "0 FR 75000 Paris ÃŽle-de-France 11 \n",
+ "1 FR 75001 Paris 01 ÃŽle-de-France 11 \n",
+ "2 FR 75001 Paris ÃŽle-de-France 11 \n",
+ "3 FR 75002 Paris 02 ÃŽle-de-France 11 \n",
+ "4 FR 75002 Paris ÃŽle-de-France 11 \n",
+ "... ... ... ... ... ... \n",
+ "51662 FR 20601 CEDEX Bastia Corse 94 \n",
+ "51663 FR 20604 CEDEX Bastia Corse 94 \n",
+ "51664 FR 20611 CEDEX Bastia Corse 94 \n",
+ "51665 FR 20620 Biguglia Corse 94 \n",
+ "51666 FR 98799 Clipperton Island NaN NaN \n",
+ "\n",
+ " admin_name2 admin_code2 admin_name3 admin_code3 latitude \\\n",
+ "0 Paris 75 NaN NaN 48.8534 \n",
+ "1 Paris 75 Paris 751 48.8592 \n",
+ "2 Paris 75 NaN NaN 48.8534 \n",
+ "3 Paris 75 Paris 751 48.8655 \n",
+ "4 Paris 75 NaN NaN 48.8534 \n",
+ "... ... ... ... ... ... \n",
+ "51662 Haute-Corse 2B Arrondissement de Bastia 2B2 42.7028 \n",
+ "51663 Haute-Corse 2B Arrondissement de Bastia 2B2 42.7028 \n",
+ "51664 Haute-Corse 2B Arrondissement de Bastia 2B2 42.7028 \n",
+ "51665 Haute-Corse 2B Arrondissement de Bastia 2B2 42.6269 \n",
+ "51666 NaN NaN NaN NaN 10.2922 \n",
+ "\n",
+ " longitude accuracy \n",
+ "0 2.3488 5.0 \n",
+ "1 2.3417 5.0 \n",
+ "2 2.3488 5.0 \n",
+ "3 2.3426 5.0 \n",
+ "4 2.3488 5.0 \n",
+ "... ... ... \n",
+ "51662 9.4500 5.0 \n",
+ "51663 9.4500 5.0 \n",
+ "51664 9.4500 5.0 \n",
+ "51665 9.4202 5.0 \n",
+ "51666 -109.2072 5.0 \n",
+ "\n",
+ "[51667 rows x 12 columns]"
+ ]
+ },
+ "execution_count": 27,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "df"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 28,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import pandas as pd\n",
+ "counts = df.placename.value_counts().reset_index().rename(columns={\"index\":\"placename\",\"placename\":\"count\"})"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 25,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "array([[<AxesSubplot:title={'center':'count'}>]], dtype=object)"
+ ]
+ },
+ "execution_count": 25,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAS+klEQVR4nO3db4xcZ3WA8efU+Ue91BRCV+CYrFNHaa24hXiUEIHQbguNE2pCUVRsWZRUbiygQW0lWhwRVemHFooUPkSkjVxIAxRlcVNKbcdVgNYrFCmFxG0gNlHAJE5jN8RAi8u4pTTm9MPcpcNmZz0zO5OZeef5SSvPfe+de8+Zax/Pnnnn3shMJEll+YlBByBJ6j2LuyQVyOIuSQWyuEtSgSzuklQgi7skFcjiLkkFsrhLPRYRRyPi9YOOQ+PN4i5JBbK4q3gRsSYiPh0R34qI70TEhyPiJyLi5oh4MiJORMTHI2JVtf10RBxbsI8fvRuPiFsiYnf1nO9FxOGIqFXrPgG8AtgbEfWI+IPnO18JLO4qXESsAPYBTwJTwGpgFri++pkBLgImgA93sOs3Vft5EbBn/rmZ+TbgX4HNmTmRmR9cfhZS5yzuKt3lwMuB38/MU5n5/cy8H9gGfCgzH8/MOnATsCUizmpzv/dn5v7MPA18AvjFvkQvdcnirtKtAZ7MzGcXjL+cxrv5eU8CZwGTbe73m02P/ws4r4P/GKS+s7irdE8Br1ik8P4bcGHT8iuAZ4FngFPAT86vqFo7L+3gmF5qVQNncVfpvgQ8DXwgIlZGxHkR8RrgbuD3ImJtREwAfwJ8qnqH/zUa78TfGBFnAzcD53ZwzGdo9PGlgbG4q2hVT3wzsI7GB53HgLcCd9LolX8BeAL4PvDu6jkngXcBHwGO03gnf2zhvpfwfuDmiPhuRLynN5lInQlv1iFJ5fGduyQVyOIuSQWyuEtSgSzuklSgofjSxfnnn59TU1MdP+/UqVOsXLmy9wENSEn5mMvwKimfknKBzvM5ePDgtzNz0e9gDEVxn5qa4qGHHur4eXNzc0xPT/c+oAEpKR9zGV4l5VNSLtB5PhHxZKt1A23LRMTmiNh18uTJQYYhScUZaHHPzL2ZuWPVqlWDDEOSiuMHqpJUIIu7JBXI4i5JBbK4S1KBLO6SVCCLuyQVaOSL+9TOewcdgiQNnZEv7pKk57K4S1KB+lLcq3tVPhQRv9qP/UuSltZWcY+IOyPiREQcWjC+KSIei4gjEbGzadV7gd29DFSS1L5237nfBWxqHoiIFcDtwNXAemBrRKyPiDcAXwVO9DBOSVIH2rrkb2Z+ISKmFgxfDhzJzMcBImIWuBaYAFbSKPj/HRH7M/OHvQtZknQmkZntbdgo7vsy89Jq+TpgU2b+VrX8NuCKzLyxWr4e+HZm7muxvx3ADoDJycmNs7OzHQdfr9d54uRpNqwu46qS9XqdiYmJQYfRE+YyvErKp6RcoPN8ZmZmDmZmbbF1fbtZR2bedYb1u4BdALVaLbu54P7c3By33n+Ko9s6f+4wKunGA+YyvErKp6RcoLf5LGe2zHFgTdPyBdVY27xZhyT1x3KK+4PAxRGxNiLOAbYAezrZgTfrkKT+aHcq5N3AA8AlEXEsIrZn5rPAjcB9wKPA7sw83MnBfecuSf3R7myZrS3G9wP7uz14Zu4F9tZqtRu63Yck6bm8/IAkFWigxd22jCT1x0CLux+oSlJ/2JaRpALZlpGkAtmWkaQC2ZaRpAJZ3CWpQPbcJalA9twlqUC2ZSSpQBZ3SSqQxV2SCuQHqpJUID9QlaQC2ZaRpAJZ3CWpQBZ3SSqQxV2SCuRsGUkqkLNlJKlAtmUkqUAWd0kqkMVdkgpkcZekAlncJalAFndJKpDz3CWpQM5zl6QC2ZaRpAJZ3CWpQBZ3SSqQxV2SCmRxl6QCWdwlqUAWd0kqkMVdkgpkcZekAvW8uEfEz0fEHRFxT0S8s9f7lySdWVvFPSLujIgTEXFowfimiHgsIo5ExE6AzHw0M98B/Drwmt6HLEk6k3bfud8FbGoeiIgVwO3A1cB6YGtErK/WvQm4F9jfs0glSW2LzGxvw4gpYF9mXlotXwnckplXVcs3AWTm+5uec29mvrHF/nYAOwAmJyc3zs7Odhx8vV7niZOn2bC6jAuP1et1JiYmBh1GT5jL8Copn5Jygc7zmZmZOZiZtcXWnbWMOFYDTzUtHwOuiIhp4C3AuSzxzj0zdwG7AGq1Wk5PT3ccwNzcHLfef4qj2zp/7jCam5ujm9dhGJnL8Copn5Jygd7ms5zivqjMnAPmer1fSVL7ljNb5jiwpmn5gmqsbb28WcfUznuXvQ9JKsVyivuDwMURsTYizgG2AHs62YE365Ck/mh3KuTdwAPAJRFxLCK2Z+azwI3AfcCjwO7MPNzJwb3NniT1R1s998zc2mJ8P8uY7piZe4G9tVrthm73IUl6Li8/IEkFGmhxty0jSf0x0OLuB6qS1B+2ZSSpQLZlJKlAtmUkqUC2ZSSpQBZ3SSqQPXdJKlBRPXcvHiZJDbZlJKlAFndJKpA9d0kqUFE9d0lSg20ZSSqQxV2SClRccXc6pCQVWNwlSc6WkaQiOVtGkgpkW0aSCmRxl6QCWdwlqUAWd0kqkMVdkgpUZHH3i0ySxl2RxV2Sxp1fYpKkAvklJkkqkG0ZSSqQxV2SCmRxl6QCFVvcnQ4paZwVW9wlaZxZ3CWpQBZ3SSqQxV2SCmRxl6QCndWPnUbEm4E3Aj8FfDQzP9uP40iSFtf2O/eIuDMiTkTEoQXjmyLisYg4EhE7ATLzM5l5A/AO4K29DVmSdCadtGXuAjY1D0TECuB24GpgPbA1ItY3bXJztX4gnOsuaVxFZra/ccQUsC8zL62WrwRuycyrquWbqk0/UP18LjM/32JfO4AdAJOTkxtnZ2c7Dr5er/PEydNsWL2KR44vfmXJDatH56Jk9XqdiYmJQYfRE+YyvErKp6RcoPN8ZmZmDmZmbbF1y+25rwaealo+BlwBvBt4PbAqItZl5h0Ln5iZu4BdALVaLaenpzs++NzcHLfef4qj26a5vsW79KPbOt/voMzNzdHN6zCMzGV4lZRPSblAb/PpyweqmXkbcNuZtouIzcDmdevW9SMMSRpby50KeRxY07R8QTXWFq/nLkn9sdzi/iBwcUSsjYhzgC3AnuWH1Tt+qCppHHUyFfJu4AHgkog4FhHbM/NZ4EbgPuBRYHdmHu5gn95mT5L6oO2ee2ZubTG+H9jfzcEzcy+wt1ar3dDN8yVJi/PyA5JUoIEWd9syktQfAy3uzpaRpP6wLSNJBbItI0kFGou2jHPdJY2bsWnLWOAljZOxKe5ggZc0Puy5S1KBxqLnLknjZqzaMpI0LizuklQge+6SVCB77pJUINsyklSgsSzuzneXVLqxLO6SVDqLuyQVyNkyklSgsZ0tY99dUslsy0hSgSzuklQgi7skFcjiLkkFGrvi7gepksbB2BV3SRoHFndJKpBfYpKkAo3tl5gkqWS2ZSSpQBZ3nEEjqTwWd0kqkMVdkgpkcZekAo11cbfXLqlUY13cJalUFndJKpDFXZIK1PPiHhEXRcRHI+KeXu9bktSetop7RNwZESci4tCC8U0R8VhEHImInQCZ+Xhmbu9HsJKk9rT7zv0uYFPzQESsAG4HrgbWA1sjYn1Po5MkdSUys70NI6aAfZl5abV8JXBLZl5VLd8EkJnvr5bvyczrltjfDmAHwOTk5MbZ2dmOg6/X6zxx8jQbVq/ikePdXVly/rkbVg/+4mX1ep2JiYlBh9ET5jK8SsqnpFyg83xmZmYOZmZtsXVnLSOO1cBTTcvHgCsi4iXAHwOvioib5ov9Qpm5C9gFUKvVcnp6uuMA5ubmuPX+UxzdNs31Xc5Zn3/u0W2dH7/X5ubm6OZ1GEbmMrxKyqekXKC3+SynuC8qM78DvKOdbSNiM7B53bp1vQ5DksbacmbLHAfWNC1fUI21zeu5S1J/LKe4PwhcHBFrI+IcYAuwpzdhSZKWo92pkHcDDwCXRMSxiNiemc8CNwL3AY8CuzPzcCcHH8bb7C3nejNeq0bSsGir556ZW1uM7wf2d3vwzNwL7K3Vajd0uw9J0nN5+QFJKtBAi/swtmUkqQQDLe7OlpGk/rAtI0kFsi3TBWfFSBp2tmUkqUC2ZSSpQLZlltCr9ottHEnPN9syklQg2zKSVCCLuyQVyOIuSQXyA9VlavVh6fPxIaof1EpqxQ9UJalAtmUkqUAWd0kqkMVdkgpkcZekAjlb5nmy2MyW5rF2Zr44O0ZSu5wtI0kFsi0jSQWyuEtSgSzuklQgi7skFcjiLkkFsrhLUoEs7pJUIL/E1KalvkDUyZeLFm57puf2+otLZ/oyVT/45Svp+eeXmCSpQLZlJKlAFndJKpDFXZIKZHGXpAJZ3CWpQBZ3SSqQxV2SCmRxl6QCWdwlqUBn9XqHEbES+DPgB8BcZn6y18eQJC2trXfuEXFnRJyIiEMLxjdFxGMRcSQidlbDbwHuycwbgDf1OF5JUhvabcvcBWxqHoiIFcDtwNXAemBrRKwHLgCeqjY73ZswJUmdiMxsb8OIKWBfZl5aLV8J3JKZV1XLN1WbHgP+IzP3RcRsZm5psb8dwA6AycnJjbOzsx0HX6/XeeLkaTasXsUjx7u7suT8czesbly8rJvH7eyn1TbN45MvgJ958aq29rlUTM3mx5tfn6Ver8W27Ua9XmdiYqLtOBfG22+dHKc5l2HS7Ws1rPl0Y1Rzaf532XwOO81nZmbmYGbWFl2ZmW39AFPAoabl64CPNC2/DfgwsBL4S+DPgW3t7Hvjxo3ZjQMHDuSF792XmZkXvndfVz/zz53XzeN29tNqm+bx2/7qM23v80zrFo4vlvNSr0fzcjcOHDjQUZztru+VTo7TnMsw6fa1GtZ8ujGquSz27zuz83yAh7JFXe35B6qZeQr4zXa2jYjNwOZ169b1OgxJGmvLmQp5HFjTtHxBNda29HruktQXyynuDwIXR8TaiDgH2ALs6U1YkqTlaHcq5N3AA8AlEXEsIrZn5rPAjcB9wKPA7sw83MnBR+k2e5I0StrquWfm1hbj+4H93R48M/cCe2u12g3d7kOS9FzeIFuSCuQNsiWpQF44TJIK1PY3VPsaRMS3gCe7eOr5wLd7HM4glZSPuQyvkvIpKRfoPJ8LM/Oli60YiuLerYh4KFt99XYElZSPuQyvkvIpKRfobT62ZSSpQBZ3SSrQqBf3XYMOoMdKysdchldJ+ZSUC/Qwn5HuuUuSFjfq79wlSYuwuEtSgUa2uLe4f+vIiIijEfFIRDwcEQ9VYy+OiM9FxNerP3960HG2sth9dVvFHw23VefqKxFx2eAif64WudwSEcer8/NwRFzTtO6mKpfHIuKqwUS9uIhYExEHIuKrEXE4In6nGh/Vc9Mqn5E7PxFxXkR8KSK+XOXyR9X42oj4YhXzp6qr7BIR51bLR6r1Ux0dsNVdPIb5B1gBfAO4CDgH+DKwftBxdZjDUeD8BWMfBHZWj3cCfzroOJeI/3XAZfz43bkWjR+4Bvh7IIBXA18cdPxt5HIL8J5Ftl1f/X07F1hb/T1cMegcmuJ7GXBZ9fiFwNeqmEf13LTKZ+TOT/UaT1SPzwa+WL3mu4Et1fgdwDurx+8C7qgebwE+1cnxRvWd++XAkcx8PDN/AMwC1w44pl64FvhY9fhjwJsHF8rSMvMLwL8vGG4V/7XAx7Phn4AXRcTLnpdA29Ail1auBWYz838y8wngCI2/j0MhM5/OzH+uHn+PxuW4VzO656ZVPq0M7fmpXuN6tXh29ZPALwH3VOMLz838ObsH+OWIiHaPN6rFfTXwVNPyMZY+4cMogc9GxMHqZuEAk5n5dPX4m8DkYELrWqv4R/V83Vi1Ku5sapGNTC7Vr/GvovEOceTPzYJ8YATPT0SsiIiHgRPA52j8ZvHdbNwfA3483h/lUq0/Cbyk3WONanEvwWsz8zLgauC3I+J1zSuz8bvYyM5THfX4adzg/WeBVwJPA7cONJoORcQE8DfA72bmfzavG8Vzs0g+I3l+MvN0Zr6Sxm1JLwd+rl/HGtXivuz7tw5aZh6v/jwB/C2NE/3M/K/E1Z8nBhdhV1rFP3LnKzOfqf4h/hD4C/7/V/uhzyUizqZRCD+ZmZ+uhkf23CyWzyifH4DM/C5wALiSRits/sZJzfH+KJdq/SrgO+0eY1SL+0jfvzUiVkbEC+cfA78CHKKRw9urzd4O/N1gIuxaq/j3AL9Rzcx4NXCyqUUwlBb0nX+NxvmBRi5bqpkMa4GLgS893/G1UvVkPwo8mpkfalo1kuemVT6jeH4i4qUR8aLq8QuAN9D4DOEAcF212cJzM3/OrgP+sfqtqz2D/gR5GZ88X0Pjk/NvAO8bdDwdxn4RjU/0vwwcno+fRj/tH4CvA58HXjzoWJfI4W4avw7/L40+4fZW8dOYJXB7da4eAWqDjr+NXD5RxfqV6h/Zy5q2f1+Vy2PA1YOOf0Eur6XRcvkK8HD1c80In5tW+Yzc+QF+AfiXKuZDwB9W4xfR+A/oCPDXwLnV+HnV8pFq/UWdHM/LD0hSgUa1LSNJWoLFXZIKZHGXpAJZ3CWpQBZ3SSqQxV2SCmRxl6QC/R/ocTcgkfkPowAAAABJRU5ErkJggg==\n",
+ "text/plain": [
+ "<Figure size 432x288 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "counts.hist(bins=500,log=True)#US"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 30,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "array([[<AxesSubplot:title={'center':'count'}>]], dtype=object)"
+ ]
+ },
+ "execution_count": 30,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 432x288 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "counts.hist(bins=500,log=True)#FR"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3.7.5 64-bit",
+ "language": "python",
+ "name": "python37564bitdc8b0e1290e74b85b0e630c435ea2fe8"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.5"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/notebooks/baseline.ipynb b/notebooks/baseline.ipynb
new file mode 100644
index 0000000..acdeb08
--- /dev/null
+++ b/notebooks/baseline.ipynb
@@ -0,0 +1,1773 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# -*- coding: utf-8 -*-\n",
+ "\n",
+ "import os\n",
+ "import argparse\n",
+ "\n",
+ "# BASIC\n",
+ "import pandas as pd\n",
+ "import numpy as np\n",
+ "\n",
+ "# ML\n",
+ "#Â MACHINE LEARNING\n",
+ "from sklearn.feature_extraction.text import CountVectorizer\n",
+ "from sklearn.feature_extraction.text import TfidfTransformer\n",
+ "from sklearn.pipeline import Pipeline\n",
+ "from sklearn.naive_bayes import MultinomialNB\n",
+ "from sklearn.svm import SVC\n",
+ "from sklearn.linear_model import SGDClassifier\n",
+ "from sklearn.tree import DecisionTreeClassifier\n",
+ "from sklearn.neighbors import KNeighborsClassifier\n",
+ "from sklearn.ensemble import RandomForestClassifier\n",
+ "\n",
+ "from sklearn.model_selection import GridSearchCV\n",
+ "\n",
+ "# ML HELPERS\n",
+ "from sklearn import preprocessing\n",
+ "from sklearn.model_selection import train_test_split\n",
+ "from sklearn.metrics import classification_report\n",
+ "\n",
+ "#PROGRESS BAR\n",
+ "from tqdm import tqdm\n",
+ "\n",
+ "import joblib\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from ngram import NGram"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "\u001b[31mFR_adjacent.csv\u001b[m\u001b[m* \u001b[31mGB_cooc.csv\u001b[m\u001b[m* TX_IDF_cooc.csv\n",
+ "\u001b[31mFR_cooc.csv\u001b[m\u001b[m* \u001b[31mGB_cooc_perm.csv\u001b[m\u001b[m* TX_IDF_inclusion.csv\n",
+ "\u001b[31mFR_inclusion.csv\u001b[m\u001b[m* \u001b[31mGB_inclusion.csv\u001b[m\u001b[m* \u001b[31mUS_adjacent.csv\u001b[m\u001b[m*\n",
+ "\u001b[31mGB_adjacent.csv\u001b[m\u001b[m* \u001b[31mGB_inclusion_perm.csv\u001b[m\u001b[m* \u001b[31mUS_cooc.csv\u001b[m\u001b[m*\n",
+ "\u001b[31mGB_adjacent_perm.csv\u001b[m\u001b[m* TX_IDF_adjacent.csv \u001b[31mUS_inclusion.csv\u001b[m\u001b[m*\n"
+ ]
+ }
+ ],
+ "source": [
+ "ls data_new_/"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 9,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "classifier_dict = {\n",
+ " \"naive-bayes\":MultinomialNB(),\n",
+ " #\"svm\":SVC(kernel=\"rbf\"),\n",
+ " \"sgd\":SGDClassifier(),\n",
+ " \"knn\":KNeighborsClassifier(),\n",
+ " \"decision-tree\": DecisionTreeClassifier(),\n",
+ " \"random-forest\":RandomForestClassifier()\n",
+ "}\n",
+ "\n",
+ "parameters = {\n",
+ " \"naive-bayes\":[{\"alpha\":[0,1]}],\n",
+ " \"svm\":[{\"kernel\":[\"rbf\",\"poly\"], 'gamma': [1e-1,1e-2,1e-3, 1,10,100]}],\n",
+ " \"sgd\":[{\"penalty\":[\"l1\",\"l2\"],\"loss\":[\"hinge\",\"modified_huber\",\"log\"]}],\n",
+ " \"knn\":[{\"n_neighbors\":list(range(4,8)),\"p\":[1,2]}],\n",
+ " \"decision-tree\": [{\"criterion\":[\"gini\",\"entropy\"]}],\n",
+ " \"random-forest\":[{\"criterion\":[\"gini\",\"entropy\"],\"n_estimators\":[10,50,100]}]\n",
+ "}\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 18,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>ID</th>\n",
+ " <th>toponym</th>\n",
+ " <th>toponym_context</th>\n",
+ " <th>latitude</th>\n",
+ " <th>longitude</th>\n",
+ " <th>hp_split</th>\n",
+ " <th>split</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>3019599</td>\n",
+ " <td>Essonne</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>48.50000</td>\n",
+ " <td>2.25000</td>\n",
+ " <td>24422</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>3013657</td>\n",
+ " <td>Hauts-de-Seine</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>48.85000</td>\n",
+ " <td>2.19293</td>\n",
+ " <td>23982</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>2975246</td>\n",
+ " <td>Seine-Saint-Denis</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>48.91421</td>\n",
+ " <td>2.47604</td>\n",
+ " <td>23983</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>2968815</td>\n",
+ " <td>Paris</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>48.85340</td>\n",
+ " <td>2.34860</td>\n",
+ " <td>23982</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>2971090</td>\n",
+ " <td>Val-de-Marne</td>\n",
+ " <td>ÃŽle-de-France</td>\n",
+ " <td>48.78149</td>\n",
+ " <td>2.49331</td>\n",
+ " <td>24423</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>...</th>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>7017</th>\n",
+ " <td>2988500</td>\n",
+ " <td>Paris Orly Airport</td>\n",
+ " <td>Paris</td>\n",
+ " <td>48.72528</td>\n",
+ " <td>2.35944</td>\n",
+ " <td>23982</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>7018</th>\n",
+ " <td>11983675</td>\n",
+ " <td>Opera Royal</td>\n",
+ " <td>Château de Versailles</td>\n",
+ " <td>48.80600</td>\n",
+ " <td>2.12290</td>\n",
+ " <td>23982</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>7019</th>\n",
+ " <td>11983678</td>\n",
+ " <td>Chapelle royale</td>\n",
+ " <td>Château de Versailles</td>\n",
+ " <td>48.80503</td>\n",
+ " <td>2.12225</td>\n",
+ " <td>23982</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>7020</th>\n",
+ " <td>6284982</td>\n",
+ " <td>Petit Trianon</td>\n",
+ " <td>Château de Versailles</td>\n",
+ " <td>48.81545</td>\n",
+ " <td>2.10976</td>\n",
+ " <td>23982</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>7021</th>\n",
+ " <td>6284981</td>\n",
+ " <td>Grand Trianon</td>\n",
+ " <td>Château de Versailles</td>\n",
+ " <td>48.81449</td>\n",
+ " <td>2.10478</td>\n",
+ " <td>23982</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "<p>7022 rows × 7 columns</p>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " ID toponym toponym_context latitude \\\n",
+ "0 3019599 Essonne ÃŽle-de-France 48.50000 \n",
+ "1 3013657 Hauts-de-Seine ÃŽle-de-France 48.85000 \n",
+ "2 2975246 Seine-Saint-Denis ÃŽle-de-France 48.91421 \n",
+ "3 2968815 Paris ÃŽle-de-France 48.85340 \n",
+ "4 2971090 Val-de-Marne ÃŽle-de-France 48.78149 \n",
+ "... ... ... ... ... \n",
+ "7017 2988500 Paris Orly Airport Paris 48.72528 \n",
+ "7018 11983675 Opera Royal Château de Versailles 48.80600 \n",
+ "7019 11983678 Chapelle royale Château de Versailles 48.80503 \n",
+ "7020 6284982 Petit Trianon Château de Versailles 48.81545 \n",
+ "7021 6284981 Grand Trianon Château de Versailles 48.81449 \n",
+ "\n",
+ " longitude hp_split split \n",
+ "0 2.25000 24422 train \n",
+ "1 2.19293 23982 train \n",
+ "2 2.47604 23983 test \n",
+ "3 2.34860 23982 train \n",
+ "4 2.49331 24423 train \n",
+ "... ... ... ... \n",
+ "7017 2.35944 23982 test \n",
+ "7018 2.12290 23982 test \n",
+ "7019 2.12225 23982 train \n",
+ "7020 2.10976 23982 train \n",
+ "7021 2.10478 23982 test \n",
+ "\n",
+ "[7022 rows x 7 columns]"
+ ]
+ },
+ "execution_count": 18,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "files = [\"TX_IDF_inclusion.csv\"]#[\"TX_IDF_adjacent.csv\",\"TX_IDF_cooc.csv\",\"TX_IDF_inclusion.csv\"]\n",
+ "basedir= \"data_new_/\"\n",
+ "df = pd.read_csv(basedir+files[0],sep=\"\\t\",index_col = 0)\n",
+ "if not len(files)<2:\n",
+ " for fn in files[1:]:\n",
+ " df = pd.concat((df,pd.read_csv(basedir + fn,sep=\"\\t\",index_col = 0)))\n",
+ "df"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 19,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "index = NGram(n=4)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 20,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "data_vectorizer = Pipeline([\n",
+ " ('vect', CountVectorizer(tokenizer=index.split)),\n",
+ " ('tfidf', TfidfTransformer()),\n",
+ "])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 21,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "X_train,y_train = (df[df.split == \"train\"].toponym + \" \" + df[df.split == \"train\"].toponym_context).values, df[df.split == \"train\"].hp_split"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 22,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "X_test,y_test = (df[df.split == \"test\"].toponym + \" \" + df[df.split == \"test\"].toponym_context).values, df[df.split == \"test\"].hp_split"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 23,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "Pipeline(steps=[('vect',\n",
+ " CountVectorizer(tokenizer=<bound method NGram.split of NGram()>)),\n",
+ " ('tfidf', TfidfTransformer())])"
+ ]
+ },
+ "execution_count": 23,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "data_vectorizer.fit((df.toponym + \" \" + df.toponym_context).values)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 24,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "X_train = data_vectorizer.transform(X_train)\n",
+ "X_test = data_vectorizer.transform(X_test)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 25,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "TRAIN AND EVAL naive-bayes\n"
+ ]
+ },
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.8/site-packages/sklearn/model_selection/_split.py:670: UserWarning: The least populated class in y has only 1 members, which is less than n_splits=5.\n",
+ " warnings.warn((\"The least populated class in y has only %d\"\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/naive_bayes.py:511: UserWarning: alpha too small will result in numeric errors, setting alpha = 1.0e-10\n",
+ " warnings.warn('alpha too small will result in numeric errors, '\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/model_selection/_split.py:670: UserWarning: The least populated class in y has only 1 members, which is less than n_splits=5.\n",
+ " warnings.warn((\"The least populated class in y has only %d\"\n"
+ ]
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Best Parameters : {'alpha': 0}\n",
+ " precision recall f1-score support\n",
+ "\n",
+ " 23545 0.00 0.00 0.00 10\n",
+ " 23546 0.50 0.13 0.21 31\n",
+ " 23547 1.00 0.17 0.29 6\n",
+ " 23981 0.00 0.00 0.00 11\n",
+ " 23982 0.62 0.96 0.75 557\n",
+ " 23983 0.74 0.51 0.61 111\n",
+ " 23984 0.00 0.00 0.00 2\n",
+ " 24422 0.62 0.35 0.45 71\n",
+ " 24423 0.70 0.59 0.64 105\n",
+ " 24424 0.33 0.07 0.11 15\n",
+ " 24867 0.80 0.40 0.53 30\n",
+ " 24868 0.00 0.00 0.00 6\n",
+ " 39537 1.00 1.00 1.00 3\n",
+ " 40045 0.00 0.00 0.00 1\n",
+ " 40047 0.00 0.00 0.00 1\n",
+ " 40558 1.00 0.50 0.67 4\n",
+ " 40559 1.00 0.50 0.67 2\n",
+ " 40561 0.00 0.00 0.00 1\n",
+ " 41070 0.00 0.00 0.00 1\n",
+ " 41071 1.00 0.75 0.86 4\n",
+ " 41072 0.00 0.00 0.00 4\n",
+ " 41073 0.67 1.00 0.80 2\n",
+ " 41583 0.85 0.58 0.69 19\n",
+ " 41584 0.00 0.00 0.00 2\n",
+ " 41585 0.00 0.00 0.00 1\n",
+ " 42094 0.00 0.00 0.00 0\n",
+ " 42095 1.00 0.67 0.80 3\n",
+ " 42096 1.00 1.00 1.00 2\n",
+ " 42097 1.00 1.00 1.00 4\n",
+ " 42606 0.00 0.00 0.00 3\n",
+ " 42607 1.00 1.00 1.00 4\n",
+ " 42608 0.00 0.00 0.00 0\n",
+ " 42609 0.00 0.00 0.00 3\n",
+ " 43117 0.00 0.00 0.00 2\n",
+ " 43118 0.50 0.50 0.50 2\n",
+ " 43119 1.00 1.00 1.00 1\n",
+ " 43120 0.00 0.00 0.00 3\n",
+ " 43121 1.00 1.00 1.00 3\n",
+ " 43122 0.00 0.00 0.00 2\n",
+ " 43124 0.00 0.00 0.00 0\n",
+ " 43630 0.00 0.00 0.00 2\n",
+ " 43631 0.67 0.50 0.57 4\n",
+ " 43632 0.00 0.00 0.00 1\n",
+ " 43633 0.00 0.00 0.00 1\n",
+ " 43634 0.00 0.00 0.00 1\n",
+ " 43635 0.00 0.00 0.00 2\n",
+ " 43636 1.00 0.55 0.71 11\n",
+ " 43637 0.00 0.00 0.00 1\n",
+ " 43638 0.00 0.00 0.00 2\n",
+ " 43639 1.00 0.33 0.50 3\n",
+ " 43640 0.00 0.00 0.00 2\n",
+ " 43641 0.00 0.00 0.00 1\n",
+ " 44142 0.67 0.67 0.67 3\n",
+ " 44143 0.73 0.80 0.76 10\n",
+ " 44144 1.00 1.00 1.00 1\n",
+ " 44146 0.00 0.00 0.00 1\n",
+ " 44147 0.00 0.00 0.00 2\n",
+ " 44148 0.00 0.00 0.00 0\n",
+ " 44149 0.00 0.00 0.00 2\n",
+ " 44150 0.69 0.56 0.62 16\n",
+ " 44151 1.00 0.29 0.44 7\n",
+ " 44152 0.75 0.75 0.75 4\n",
+ " 44153 0.00 0.00 0.00 1\n",
+ " 44154 0.00 0.00 0.00 1\n",
+ " 44654 0.00 0.00 0.00 1\n",
+ " 44655 1.00 0.50 0.67 4\n",
+ " 44656 0.00 0.00 0.00 2\n",
+ " 44657 0.00 0.00 0.00 2\n",
+ " 44658 1.00 0.50 0.67 2\n",
+ " 44660 0.00 0.00 0.00 4\n",
+ " 44661 1.00 0.67 0.80 9\n",
+ " 44662 0.57 0.72 0.64 64\n",
+ " 44663 0.50 0.50 0.50 12\n",
+ " 44664 0.50 0.67 0.57 3\n",
+ " 44665 1.00 0.67 0.80 3\n",
+ " 44666 1.00 0.33 0.50 3\n",
+ " 45165 0.50 1.00 0.67 1\n",
+ " 45168 1.00 1.00 1.00 2\n",
+ " 45169 0.50 1.00 0.67 1\n",
+ " 45170 0.67 1.00 0.80 2\n",
+ " 45171 0.00 0.00 0.00 2\n",
+ " 45172 1.00 0.56 0.71 9\n",
+ " 45173 0.58 0.84 0.69 57\n",
+ " 45174 0.50 0.86 0.63 99\n",
+ " 45175 0.60 0.43 0.50 7\n",
+ " 45176 0.75 1.00 0.86 3\n",
+ " 45177 1.00 0.50 0.67 6\n",
+ " 45678 0.00 0.00 0.00 3\n",
+ " 45679 1.00 0.29 0.44 7\n",
+ " 45680 0.00 0.00 0.00 1\n",
+ " 45681 0.00 0.00 0.00 1\n",
+ " 45682 0.80 0.50 0.62 8\n",
+ " 45683 1.00 0.25 0.40 4\n",
+ " 45684 0.00 0.00 0.00 0\n",
+ " 45685 0.80 0.73 0.76 11\n",
+ " 45686 0.67 0.43 0.53 23\n",
+ " 45687 0.73 0.67 0.70 12\n",
+ " 45688 0.50 0.29 0.36 7\n",
+ " 45689 0.71 0.94 0.81 16\n",
+ " 45690 1.00 0.40 0.57 5\n",
+ " 46184 0.00 0.00 0.00 3\n",
+ " 46189 0.00 0.00 0.00 1\n",
+ " 46190 0.83 0.62 0.71 8\n",
+ " 46191 0.00 0.00 0.00 1\n",
+ " 46192 0.00 0.00 0.00 1\n",
+ " 46193 1.00 1.00 1.00 2\n",
+ " 46194 0.33 1.00 0.50 1\n",
+ " 46195 0.00 0.00 0.00 4\n",
+ " 46196 0.33 0.33 0.33 3\n",
+ " 46197 0.25 0.14 0.18 7\n",
+ " 46198 0.33 0.20 0.25 5\n",
+ " 46199 0.50 0.20 0.29 5\n",
+ " 46200 0.83 0.31 0.45 16\n",
+ " 46201 1.00 0.56 0.71 9\n",
+ " 46697 0.71 0.38 0.50 13\n",
+ " 46701 0.50 0.50 0.50 2\n",
+ " 46702 0.00 0.00 0.00 1\n",
+ " 46703 0.00 0.00 0.00 1\n",
+ " 46707 1.00 0.25 0.40 4\n",
+ " 46709 1.00 0.17 0.29 6\n",
+ " 46710 0.29 0.14 0.19 14\n",
+ " 46711 0.40 0.33 0.36 6\n",
+ " 46712 0.00 0.00 0.00 4\n",
+ " 46713 0.60 1.00 0.75 3\n",
+ " 46714 0.00 0.00 0.00 1\n",
+ " 47212 0.33 0.50 0.40 2\n",
+ " 47213 0.00 0.00 0.00 1\n",
+ " 47214 0.00 0.00 0.00 1\n",
+ " 47217 1.00 0.25 0.40 4\n",
+ " 47218 1.00 1.00 1.00 1\n",
+ " 47219 0.00 0.00 0.00 2\n",
+ " 47220 1.00 0.75 0.86 4\n",
+ " 47221 0.40 0.62 0.48 13\n",
+ " 47222 1.00 0.33 0.50 6\n",
+ " 47223 1.00 0.50 0.67 4\n",
+ " 47224 0.50 1.00 0.67 1\n",
+ " 47225 1.00 0.67 0.80 9\n",
+ " 47226 1.00 1.00 1.00 1\n",
+ " 47723 0.00 0.00 0.00 2\n",
+ " 47726 0.00 0.00 0.00 1\n",
+ " 47730 1.00 1.00 1.00 1\n",
+ " 47733 0.75 0.50 0.60 12\n",
+ " 47734 1.00 0.50 0.67 4\n",
+ " 47735 0.33 0.50 0.40 2\n",
+ " 47736 0.00 0.00 0.00 1\n",
+ " 47737 0.50 0.50 0.50 2\n",
+ " 47738 0.33 0.33 0.33 3\n",
+ " 48235 0.00 0.00 0.00 1\n",
+ " 48236 0.00 0.00 0.00 1\n",
+ " 48237 1.00 0.50 0.67 2\n",
+ " 48239 0.00 0.00 0.00 1\n",
+ " 48240 1.00 1.00 1.00 1\n",
+ " 48243 0.00 0.00 0.00 1\n",
+ " 48244 0.80 0.67 0.73 6\n",
+ " 48245 0.75 0.30 0.43 10\n",
+ " 48246 0.50 0.40 0.44 5\n",
+ " 48247 0.50 0.50 0.50 2\n",
+ " 48248 0.50 0.33 0.40 3\n",
+ " 48249 0.33 0.50 0.40 2\n",
+ " 48250 0.00 0.00 0.00 3\n",
+ " 48748 0.00 0.00 0.00 1\n",
+ " 48749 0.00 0.00 0.00 1\n",
+ " 48753 0.00 0.00 0.00 3\n",
+ " 48754 1.00 0.50 0.67 4\n",
+ " 48755 1.00 1.00 1.00 1\n",
+ " 48756 1.00 0.67 0.80 3\n",
+ " 48757 0.66 0.75 0.70 68\n",
+ " 48758 0.00 0.00 0.00 2\n",
+ " 48759 0.56 0.42 0.48 12\n",
+ " 48760 0.88 0.44 0.58 16\n",
+ " 48761 0.60 0.43 0.50 7\n",
+ " 48762 0.00 0.00 0.00 7\n",
+ " 49265 0.00 0.00 0.00 1\n",
+ " 49266 0.00 0.00 0.00 3\n",
+ " 49267 1.00 0.50 0.67 2\n",
+ " 49268 0.75 0.38 0.50 16\n",
+ " 49269 0.67 0.40 0.50 5\n",
+ " 49270 0.50 0.33 0.40 3\n",
+ " 49271 0.00 0.00 0.00 2\n",
+ " 49272 0.52 0.62 0.57 40\n",
+ " 49273 0.00 0.00 0.00 2\n",
+ " 49274 0.71 0.45 0.56 11\n",
+ " 49773 0.00 0.00 0.00 0\n",
+ " 49776 0.00 0.00 0.00 1\n",
+ " 49778 0.00 0.00 0.00 1\n",
+ " 49779 1.00 0.33 0.50 3\n",
+ " 49780 0.60 0.86 0.71 70\n",
+ " 49781 0.00 0.00 0.00 4\n",
+ " 49782 0.50 0.25 0.33 4\n",
+ " 49783 0.00 0.00 0.00 0\n",
+ " 49784 0.76 0.63 0.69 65\n",
+ " 49785 0.59 0.82 0.68 33\n",
+ " 50283 0.00 0.00 0.00 1\n",
+ " 50288 0.00 0.00 0.00 2\n",
+ " 50290 0.00 0.00 0.00 1\n",
+ " 50291 0.38 0.45 0.42 11\n",
+ " 50292 0.80 0.57 0.67 7\n",
+ " 50293 0.67 1.00 0.80 2\n",
+ " 50295 0.00 0.00 0.00 3\n",
+ " 50296 0.44 0.40 0.42 10\n",
+ " 50297 1.00 0.30 0.46 10\n",
+ " 50801 1.00 0.25 0.40 4\n",
+ " 50802 0.00 0.00 0.00 2\n",
+ " 50804 1.00 0.50 0.67 2\n",
+ " 50805 1.00 1.00 1.00 3\n",
+ " 50806 0.00 0.00 0.00 2\n",
+ " 50807 0.00 0.00 0.00 3\n",
+ " 50808 0.25 0.17 0.20 6\n",
+ " 50809 0.00 0.00 0.00 0\n",
+ " 51313 0.00 0.00 0.00 0\n",
+ " 51315 0.00 0.00 0.00 1\n",
+ " 51316 1.00 0.50 0.67 2\n",
+ " 51317 0.00 0.00 0.00 1\n",
+ " 51318 1.00 0.50 0.67 2\n",
+ " 51320 0.50 1.00 0.67 1\n",
+ " 51827 0.00 0.00 0.00 1\n",
+ " 51828 0.00 0.00 0.00 1\n",
+ " 51829 0.00 0.00 0.00 1\n",
+ " 51830 0.80 1.00 0.89 4\n",
+ " 52340 0.40 0.80 0.53 5\n",
+ " 52341 0.89 0.73 0.80 22\n",
+ " 52851 0.00 0.00 0.00 1\n",
+ " 52852 1.00 0.33 0.50 3\n",
+ " 52853 0.83 0.71 0.77 7\n",
+ " 52854 0.00 0.00 0.00 4\n",
+ " 53362 1.00 0.25 0.40 8\n",
+ " 53364 0.00 0.00 0.00 1\n",
+ " 53875 0.00 0.00 0.00 2\n",
+ " 54387 0.00 0.00 0.00 0\n",
+ " 54388 0.00 0.00 0.00 1\n",
+ " 54389 0.00 0.00 0.00 0\n",
+ " 54899 0.00 0.00 0.00 1\n",
+ " 54900 0.78 0.41 0.54 17\n",
+ " 54901 0.67 0.36 0.47 11\n",
+ " 54902 1.00 0.88 0.93 8\n",
+ " 55412 0.33 1.00 0.50 1\n",
+ " 55413 1.00 0.44 0.62 9\n",
+ "\n",
+ " accuracy 0.63 2318\n",
+ " macro avg 0.44 0.34 0.36 2318\n",
+ "weighted avg 0.62 0.63 0.59 2318\n",
+ "\n",
+ "TRAIN AND EVAL sgd\n",
+ "Best Parameters : {'loss': 'hinge', 'penalty': 'l2'}\n",
+ " precision recall f1-score support\n",
+ "\n",
+ " 23545 1.00 0.10 0.18 10\n",
+ " 23546 0.62 0.81 0.70 31\n",
+ " 23547 0.00 0.00 0.00 6\n",
+ " 23981 0.00 0.00 0.00 11\n",
+ " 23982 0.91 0.96 0.93 557\n",
+ " 23983 0.81 0.77 0.79 111\n",
+ " 23984 0.00 0.00 0.00 2\n",
+ " 24422 0.79 0.75 0.77 71\n",
+ " 24423 0.71 0.70 0.70 105\n",
+ " 24424 0.33 0.27 0.30 15\n",
+ " 24867 0.76 0.87 0.81 30\n",
+ " 24868 1.00 0.17 0.29 6\n",
+ " 39537 1.00 1.00 1.00 3\n",
+ " 40045 0.00 0.00 0.00 1\n",
+ " 40047 0.00 0.00 0.00 1\n",
+ " 40558 0.60 0.75 0.67 4\n",
+ " 40559 0.67 1.00 0.80 2\n",
+ " 40561 0.00 0.00 0.00 1\n",
+ " 41070 0.00 0.00 0.00 1\n",
+ " 41071 0.75 0.75 0.75 4\n",
+ " 41072 1.00 0.25 0.40 4\n",
+ " 41073 0.33 1.00 0.50 2\n",
+ " 41582 0.00 0.00 0.00 0\n",
+ " 41583 0.95 1.00 0.97 19\n",
+ " 41584 1.00 0.50 0.67 2\n",
+ " 41585 0.00 0.00 0.00 1\n",
+ " 42095 0.67 0.67 0.67 3\n",
+ " 42096 1.00 1.00 1.00 2\n",
+ " 42097 1.00 1.00 1.00 4\n",
+ " 42606 0.00 0.00 0.00 3\n",
+ " 42607 1.00 1.00 1.00 4\n",
+ " 42608 0.00 0.00 0.00 0\n",
+ " 42609 0.00 0.00 0.00 3\n",
+ " 43117 0.00 0.00 0.00 2\n",
+ " 43118 0.67 1.00 0.80 2\n",
+ " 43119 1.00 1.00 1.00 1\n",
+ " 43120 1.00 0.33 0.50 3\n",
+ " 43121 1.00 1.00 1.00 3\n",
+ " 43122 1.00 1.00 1.00 2\n",
+ " 43630 0.00 0.00 0.00 2\n",
+ " 43631 0.50 0.25 0.33 4\n",
+ " 43632 0.00 0.00 0.00 1\n",
+ " 43633 0.00 0.00 0.00 1\n",
+ " 43634 1.00 1.00 1.00 1\n",
+ " 43635 0.00 0.00 0.00 2\n",
+ " 43636 1.00 0.64 0.78 11\n",
+ " 43637 0.50 1.00 0.67 1\n",
+ " 43638 1.00 0.50 0.67 2\n",
+ " 43639 0.00 0.00 0.00 3\n",
+ " 43640 0.50 0.50 0.50 2\n",
+ " 43641 0.00 0.00 0.00 1\n",
+ " 44141 0.00 0.00 0.00 0\n",
+ " 44142 0.67 0.67 0.67 3\n",
+ " 44143 0.73 0.80 0.76 10\n",
+ " 44144 0.33 1.00 0.50 1\n",
+ " 44146 0.00 0.00 0.00 1\n",
+ " 44147 0.00 0.00 0.00 2\n",
+ " 44148 0.00 0.00 0.00 0\n",
+ " 44149 0.00 0.00 0.00 2\n",
+ " 44150 0.75 0.75 0.75 16\n",
+ " 44151 0.50 0.57 0.53 7\n",
+ " 44152 0.80 1.00 0.89 4\n",
+ " 44153 0.00 0.00 0.00 1\n",
+ " 44154 0.00 0.00 0.00 1\n",
+ " 44654 0.00 0.00 0.00 1\n",
+ " 44655 1.00 1.00 1.00 4\n",
+ " 44656 1.00 0.50 0.67 2\n",
+ " 44657 0.00 0.00 0.00 2\n",
+ " 44658 1.00 1.00 1.00 2\n",
+ " 44660 0.50 0.50 0.50 4\n",
+ " 44661 1.00 0.89 0.94 9\n",
+ " 44662 0.80 0.86 0.83 64\n",
+ " 44663 0.53 0.75 0.62 12\n",
+ " 44664 0.67 0.67 0.67 3\n",
+ " 44665 0.40 0.67 0.50 3\n",
+ " 44666 0.50 0.67 0.57 3\n",
+ " 45165 1.00 1.00 1.00 1\n",
+ " 45168 1.00 1.00 1.00 2\n",
+ " 45169 0.50 1.00 0.67 1\n",
+ " 45170 0.67 1.00 0.80 2\n",
+ " 45171 1.00 0.50 0.67 2\n",
+ " 45172 0.83 0.56 0.67 9\n",
+ " 45173 0.76 0.88 0.81 57\n",
+ " 45174 0.86 0.89 0.88 99\n",
+ " 45175 0.75 0.43 0.55 7\n",
+ " 45176 1.00 1.00 1.00 3\n",
+ " 45177 0.50 0.17 0.25 6\n",
+ " 45678 0.00 0.00 0.00 3\n",
+ " 45679 1.00 0.43 0.60 7\n",
+ " 45680 0.00 0.00 0.00 1\n",
+ " 45681 0.00 0.00 0.00 1\n",
+ " 45682 0.75 0.75 0.75 8\n",
+ " 45683 0.00 0.00 0.00 4\n",
+ " 45684 0.00 0.00 0.00 0\n",
+ " 45685 0.64 0.82 0.72 11\n",
+ " 45686 0.71 0.65 0.68 23\n",
+ " 45687 0.92 0.92 0.92 12\n",
+ " 45688 0.67 0.57 0.62 7\n",
+ " 45689 0.68 0.94 0.79 16\n",
+ " 45690 1.00 0.60 0.75 5\n",
+ " 46184 0.00 0.00 0.00 3\n",
+ " 46189 0.00 0.00 0.00 1\n",
+ " 46190 0.67 1.00 0.80 8\n",
+ " 46191 0.00 0.00 0.00 1\n",
+ " 46192 0.00 0.00 0.00 1\n",
+ " 46193 1.00 1.00 1.00 2\n",
+ " 46194 0.33 1.00 0.50 1\n",
+ " 46195 1.00 0.75 0.86 4\n",
+ " 46196 0.33 0.67 0.44 3\n",
+ " 46197 0.50 0.57 0.53 7\n",
+ " 46198 0.71 1.00 0.83 5\n",
+ " 46199 0.62 1.00 0.77 5\n",
+ " 46200 0.79 0.69 0.73 16\n",
+ " 46201 1.00 0.44 0.62 9\n",
+ " 46697 0.81 1.00 0.90 13\n",
+ " 46701 1.00 0.50 0.67 2\n",
+ " 46702 1.00 1.00 1.00 1\n",
+ " 46703 0.00 0.00 0.00 1\n",
+ " 46707 1.00 0.75 0.86 4\n",
+ " 46709 0.67 0.33 0.44 6\n",
+ " 46710 0.71 0.36 0.48 14\n",
+ " 46711 0.83 0.83 0.83 6\n",
+ " 46712 1.00 0.75 0.86 4\n",
+ " 46713 1.00 0.67 0.80 3\n",
+ " 46714 0.50 1.00 0.67 1\n",
+ " 47212 0.50 1.00 0.67 2\n",
+ " 47213 0.33 1.00 0.50 1\n",
+ " 47214 0.00 0.00 0.00 1\n",
+ " 47217 1.00 0.75 0.86 4\n",
+ " 47218 1.00 1.00 1.00 1\n",
+ " 47219 0.50 0.50 0.50 2\n",
+ " 47220 1.00 0.75 0.86 4\n",
+ " 47221 0.38 0.69 0.49 13\n",
+ " 47222 1.00 0.17 0.29 6\n",
+ " 47223 0.60 0.75 0.67 4\n",
+ " 47224 0.33 1.00 0.50 1\n",
+ " 47225 1.00 1.00 1.00 9\n",
+ " 47226 1.00 1.00 1.00 1\n",
+ " 47723 1.00 1.00 1.00 2\n",
+ " 47726 0.00 0.00 0.00 1\n",
+ " 47727 0.00 0.00 0.00 0\n",
+ " 47730 1.00 1.00 1.00 1\n",
+ " 47733 0.75 0.75 0.75 12\n",
+ " 47734 0.50 0.50 0.50 4\n",
+ " 47735 0.25 0.50 0.33 2\n",
+ " 47736 0.00 0.00 0.00 1\n",
+ " 47737 0.50 0.50 0.50 2\n",
+ " 47738 0.25 0.33 0.29 3\n",
+ " 48235 1.00 1.00 1.00 1\n",
+ " 48236 0.00 0.00 0.00 1\n",
+ " 48237 1.00 1.00 1.00 2\n",
+ " 48239 0.00 0.00 0.00 1\n",
+ " 48240 1.00 1.00 1.00 1\n",
+ " 48243 0.00 0.00 0.00 1\n",
+ " 48244 1.00 0.67 0.80 6\n",
+ " 48245 0.57 0.40 0.47 10\n",
+ " 48246 0.50 0.40 0.44 5\n",
+ " 48247 1.00 0.50 0.67 2\n",
+ " 48248 0.67 0.67 0.67 3\n",
+ " 48249 0.20 0.50 0.29 2\n",
+ " 48250 1.00 0.33 0.50 3\n",
+ " 48748 0.00 0.00 0.00 1\n",
+ " 48749 1.00 1.00 1.00 1\n",
+ " 48753 1.00 1.00 1.00 3\n",
+ " 48754 1.00 1.00 1.00 4\n",
+ " 48755 0.20 1.00 0.33 1\n",
+ " 48756 0.50 0.67 0.57 3\n",
+ " 48757 0.89 0.99 0.94 68\n",
+ " 48758 0.00 0.00 0.00 2\n",
+ " 48759 0.85 0.92 0.88 12\n",
+ " 48760 0.62 0.50 0.55 16\n",
+ " 48761 1.00 0.57 0.73 7\n",
+ " 48762 0.00 0.00 0.00 7\n",
+ " 49259 0.00 0.00 0.00 0\n",
+ " 49265 0.00 0.00 0.00 1\n",
+ " 49266 0.00 0.00 0.00 3\n",
+ " 49267 1.00 0.50 0.67 2\n",
+ " 49268 0.94 0.94 0.94 16\n",
+ " 49269 0.67 0.40 0.50 5\n",
+ " 49270 0.50 0.33 0.40 3\n",
+ " 49271 1.00 0.50 0.67 2\n",
+ " 49272 0.59 0.80 0.68 40\n",
+ " 49273 0.11 0.50 0.18 2\n",
+ " 49274 0.75 0.82 0.78 11\n",
+ " 49773 0.00 0.00 0.00 0\n",
+ " 49776 0.00 0.00 0.00 1\n",
+ " 49778 0.50 1.00 0.67 1\n",
+ " 49779 1.00 0.33 0.50 3\n",
+ " 49780 0.85 1.00 0.92 70\n",
+ " 49781 1.00 0.75 0.86 4\n",
+ " 49782 1.00 0.25 0.40 4\n",
+ " 49783 0.00 0.00 0.00 0\n",
+ " 49784 0.88 0.71 0.79 65\n",
+ " 49785 0.74 0.97 0.84 33\n",
+ " 50283 0.00 0.00 0.00 1\n",
+ " 50288 0.50 1.00 0.67 2\n",
+ " 50289 0.00 0.00 0.00 0\n",
+ " 50290 0.00 0.00 0.00 1\n",
+ " 50291 0.75 0.55 0.63 11\n",
+ " 50292 0.83 0.71 0.77 7\n",
+ " 50293 0.50 0.50 0.50 2\n",
+ " 50295 0.00 0.00 0.00 3\n",
+ " 50296 0.50 0.50 0.50 10\n",
+ " 50297 1.00 0.90 0.95 10\n",
+ " 50801 1.00 0.50 0.67 4\n",
+ " 50802 0.50 0.50 0.50 2\n",
+ " 50803 0.00 0.00 0.00 0\n",
+ " 50804 0.67 1.00 0.80 2\n",
+ " 50805 0.67 0.67 0.67 3\n",
+ " 50806 1.00 1.00 1.00 2\n",
+ " 50807 0.00 0.00 0.00 3\n",
+ " 50808 0.40 0.33 0.36 6\n",
+ " 50809 0.00 0.00 0.00 0\n",
+ " 51313 0.00 0.00 0.00 0\n",
+ " 51314 0.00 0.00 0.00 0\n",
+ " 51315 0.00 0.00 0.00 1\n",
+ " 51316 1.00 1.00 1.00 2\n",
+ " 51317 1.00 1.00 1.00 1\n",
+ " 51318 1.00 1.00 1.00 2\n",
+ " 51319 0.00 0.00 0.00 0\n",
+ " 51320 0.00 0.00 0.00 1\n",
+ " 51827 1.00 1.00 1.00 1\n",
+ " 51828 0.00 0.00 0.00 1\n",
+ " 51829 0.33 1.00 0.50 1\n",
+ " 51830 0.75 0.75 0.75 4\n",
+ " 52340 0.44 0.80 0.57 5\n",
+ " 52341 0.86 0.82 0.84 22\n",
+ " 52851 0.00 0.00 0.00 1\n",
+ " 52852 1.00 0.67 0.80 3\n",
+ " 52853 1.00 0.71 0.83 7\n",
+ " 52854 1.00 0.50 0.67 4\n",
+ " 53362 0.80 1.00 0.89 8\n",
+ " 53364 0.00 0.00 0.00 1\n",
+ " 53875 0.00 0.00 0.00 2\n",
+ " 54387 0.00 0.00 0.00 0\n",
+ " 54388 0.00 0.00 0.00 1\n",
+ " 54899 0.00 0.00 0.00 1\n",
+ " 54900 0.80 0.71 0.75 17\n",
+ " 54901 0.82 0.82 0.82 11\n",
+ " 54902 1.00 1.00 1.00 8\n",
+ " 55412 0.25 1.00 0.40 1\n",
+ " 55413 1.00 0.78 0.88 9\n",
+ "\n",
+ " accuracy 0.78 2318\n",
+ " macro avg 0.54 0.52 0.50 2318\n",
+ "weighted avg 0.78 0.78 0.77 2318\n",
+ "\n",
+ "TRAIN AND EVAL knn\n"
+ ]
+ },
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/model_selection/_split.py:670: UserWarning: The least populated class in y has only 1 members, which is less than n_splits=5.\n",
+ " warnings.warn((\"The least populated class in y has only %d\"\n"
+ ]
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Best Parameters : {'n_neighbors': 4, 'p': 2}\n"
+ ]
+ },
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/model_selection/_split.py:670: UserWarning: The least populated class in y has only 1 members, which is less than n_splits=5.\n",
+ " warnings.warn((\"The least populated class in y has only %d\"\n"
+ ]
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ " precision recall f1-score support\n",
+ "\n",
+ " 23545 0.00 0.00 0.00 10\n",
+ " 23546 0.54 0.81 0.65 31\n",
+ " 23547 0.00 0.00 0.00 6\n",
+ " 23981 0.00 0.00 0.00 11\n",
+ " 23982 0.86 0.95 0.90 557\n",
+ " 23983 0.81 0.75 0.78 111\n",
+ " 23984 0.00 0.00 0.00 2\n",
+ " 24422 0.80 0.72 0.76 71\n",
+ " 24423 0.71 0.60 0.65 105\n",
+ " 24424 0.32 0.47 0.38 15\n",
+ " 24867 0.76 0.73 0.75 30\n",
+ " 24868 0.00 0.00 0.00 6\n",
+ " 39537 1.00 0.33 0.50 3\n",
+ " 40045 0.00 0.00 0.00 1\n",
+ " 40046 0.00 0.00 0.00 0\n",
+ " 40047 0.00 0.00 0.00 1\n",
+ " 40049 0.00 0.00 0.00 0\n",
+ " 40558 0.60 0.75 0.67 4\n",
+ " 40559 0.25 0.50 0.33 2\n",
+ " 40560 0.00 0.00 0.00 0\n",
+ " 40561 0.00 0.00 0.00 1\n",
+ " 41070 0.00 0.00 0.00 1\n",
+ " 41071 0.50 0.75 0.60 4\n",
+ " 41072 1.00 0.50 0.67 4\n",
+ " 41073 0.17 0.50 0.25 2\n",
+ " 41583 0.82 0.95 0.88 19\n",
+ " 41584 0.20 0.50 0.29 2\n",
+ " 41585 0.00 0.00 0.00 1\n",
+ " 42094 0.00 0.00 0.00 0\n",
+ " 42095 0.33 0.33 0.33 3\n",
+ " 42096 1.00 1.00 1.00 2\n",
+ " 42097 1.00 0.75 0.86 4\n",
+ " 42606 0.00 0.00 0.00 3\n",
+ " 42607 1.00 1.00 1.00 4\n",
+ " 42608 0.00 0.00 0.00 0\n",
+ " 42609 0.08 0.33 0.13 3\n",
+ " 43117 0.00 0.00 0.00 2\n",
+ " 43118 0.40 1.00 0.57 2\n",
+ " 43119 1.00 1.00 1.00 1\n",
+ " 43120 0.00 0.00 0.00 3\n",
+ " 43121 1.00 0.67 0.80 3\n",
+ " 43122 1.00 1.00 1.00 2\n",
+ " 43123 0.00 0.00 0.00 0\n",
+ " 43124 0.00 0.00 0.00 0\n",
+ " 43630 0.00 0.00 0.00 2\n",
+ " 43631 0.29 0.50 0.36 4\n",
+ " 43632 0.00 0.00 0.00 1\n",
+ " 43633 0.00 0.00 0.00 1\n",
+ " 43634 0.50 1.00 0.67 1\n",
+ " 43635 0.00 0.00 0.00 2\n",
+ " 43636 0.83 0.45 0.59 11\n",
+ " 43637 0.50 1.00 0.67 1\n",
+ " 43638 0.50 0.50 0.50 2\n",
+ " 43639 0.00 0.00 0.00 3\n",
+ " 43640 0.14 0.50 0.22 2\n",
+ " 43641 0.00 0.00 0.00 1\n",
+ " 44141 0.00 0.00 0.00 0\n",
+ " 44142 0.67 0.67 0.67 3\n",
+ " 44143 0.53 0.80 0.64 10\n",
+ " 44144 0.50 1.00 0.67 1\n",
+ " 44146 0.00 0.00 0.00 1\n",
+ " 44147 0.00 0.00 0.00 2\n",
+ " 44148 0.00 0.00 0.00 0\n",
+ " 44149 0.00 0.00 0.00 2\n",
+ " 44150 0.69 0.69 0.69 16\n",
+ " 44151 0.56 0.71 0.63 7\n",
+ " 44152 0.75 0.75 0.75 4\n",
+ " 44153 0.00 0.00 0.00 1\n",
+ " 44154 0.00 0.00 0.00 1\n",
+ " 44654 0.00 0.00 0.00 1\n",
+ " 44655 0.75 0.75 0.75 4\n",
+ " 44656 0.00 0.00 0.00 2\n",
+ " 44657 0.00 0.00 0.00 2\n",
+ " 44658 1.00 1.00 1.00 2\n",
+ " 44660 0.00 0.00 0.00 4\n",
+ " 44661 0.47 0.78 0.58 9\n",
+ " 44662 0.64 0.91 0.75 64\n",
+ " 44663 0.38 0.42 0.40 12\n",
+ " 44664 0.33 0.33 0.33 3\n",
+ " 44665 0.33 0.33 0.33 3\n",
+ " 44666 0.25 0.67 0.36 3\n",
+ " 45165 1.00 1.00 1.00 1\n",
+ " 45168 1.00 0.50 0.67 2\n",
+ " 45169 0.50 1.00 0.67 1\n",
+ " 45170 1.00 0.50 0.67 2\n",
+ " 45171 0.00 0.00 0.00 2\n",
+ " 45172 0.86 0.67 0.75 9\n",
+ " 45173 0.71 0.84 0.77 57\n",
+ " 45174 0.67 0.81 0.73 99\n",
+ " 45175 0.00 0.00 0.00 7\n",
+ " 45176 0.75 1.00 0.86 3\n",
+ " 45177 0.00 0.00 0.00 6\n",
+ " 45678 0.00 0.00 0.00 3\n",
+ " 45679 1.00 0.14 0.25 7\n",
+ " 45680 0.00 0.00 0.00 1\n",
+ " 45681 0.00 0.00 0.00 1\n",
+ " 45682 0.67 0.75 0.71 8\n",
+ " 45683 1.00 0.25 0.40 4\n",
+ " 45684 0.00 0.00 0.00 0\n",
+ " 45685 0.57 0.36 0.44 11\n",
+ " 45686 0.59 0.43 0.50 23\n",
+ " 45687 0.75 0.50 0.60 12\n",
+ " 45688 0.38 0.43 0.40 7\n",
+ " 45689 0.68 0.94 0.79 16\n",
+ " 45690 1.00 0.60 0.75 5\n",
+ " 46184 0.20 0.67 0.31 3\n",
+ " 46189 0.00 0.00 0.00 1\n",
+ " 46190 0.70 0.88 0.78 8\n",
+ " 46191 0.00 0.00 0.00 1\n",
+ " 46192 0.00 0.00 0.00 1\n",
+ " 46193 0.67 1.00 0.80 2\n",
+ " 46194 0.50 1.00 0.67 1\n",
+ " 46195 0.75 0.75 0.75 4\n",
+ " 46196 0.40 0.67 0.50 3\n",
+ " 46197 0.17 0.14 0.15 7\n",
+ " 46198 0.80 0.80 0.80 5\n",
+ " 46199 0.00 0.00 0.00 5\n",
+ " 46200 0.62 0.50 0.55 16\n",
+ " 46201 0.43 0.33 0.38 9\n",
+ " 46697 0.86 0.46 0.60 13\n",
+ " 46701 1.00 1.00 1.00 2\n",
+ " 46702 0.00 0.00 0.00 1\n",
+ " 46703 0.00 0.00 0.00 1\n",
+ " 46705 0.00 0.00 0.00 0\n",
+ " 46707 0.67 0.50 0.57 4\n",
+ " 46708 0.00 0.00 0.00 0\n",
+ " 46709 0.50 0.33 0.40 6\n",
+ " 46710 0.57 0.57 0.57 14\n",
+ " 46711 0.71 0.83 0.77 6\n",
+ " 46712 0.00 0.00 0.00 4\n",
+ " 46713 1.00 0.33 0.50 3\n",
+ " 46714 1.00 1.00 1.00 1\n",
+ " 47212 0.67 1.00 0.80 2\n",
+ " 47213 0.50 1.00 0.67 1\n",
+ " 47214 0.00 0.00 0.00 1\n",
+ " 47217 1.00 0.50 0.67 4\n",
+ " 47218 1.00 1.00 1.00 1\n",
+ " 47219 0.50 0.50 0.50 2\n",
+ " 47220 0.50 0.50 0.50 4\n",
+ " 47221 0.70 0.54 0.61 13\n",
+ " 47222 1.00 0.17 0.29 6\n",
+ " 47223 0.67 0.50 0.57 4\n",
+ " 47224 0.00 0.00 0.00 1\n",
+ " 47225 0.90 1.00 0.95 9\n",
+ " 47226 0.50 1.00 0.67 1\n",
+ " 47723 1.00 1.00 1.00 2\n",
+ " 47726 0.00 0.00 0.00 1\n",
+ " 47730 0.00 0.00 0.00 1\n",
+ " 47733 0.75 0.75 0.75 12\n",
+ " 47734 0.40 0.50 0.44 4\n",
+ " 47735 0.25 0.50 0.33 2\n",
+ " 47736 0.00 0.00 0.00 1\n",
+ " 47737 0.00 0.00 0.00 2\n",
+ " 47738 0.33 0.33 0.33 3\n",
+ " 48235 0.00 0.00 0.00 1\n",
+ " 48236 0.00 0.00 0.00 1\n",
+ " 48237 1.00 1.00 1.00 2\n",
+ " 48239 0.00 0.00 0.00 1\n",
+ " 48240 1.00 1.00 1.00 1\n",
+ " 48243 0.00 0.00 0.00 1\n",
+ " 48244 0.75 0.50 0.60 6\n",
+ " 48245 0.50 0.40 0.44 10\n",
+ " 48246 0.00 0.00 0.00 5\n",
+ " 48247 0.00 0.00 0.00 2\n",
+ " 48248 1.00 0.67 0.80 3\n",
+ " 48249 0.00 0.00 0.00 2\n",
+ " 48250 0.00 0.00 0.00 3\n",
+ " 48748 0.00 0.00 0.00 1\n",
+ " 48749 0.00 0.00 0.00 1\n",
+ " 48753 1.00 1.00 1.00 3\n",
+ " 48754 1.00 0.50 0.67 4\n",
+ " 48755 0.33 1.00 0.50 1\n",
+ " 48756 0.50 0.67 0.57 3\n",
+ " 48757 0.94 0.94 0.94 68\n",
+ " 48758 0.00 0.00 0.00 2\n",
+ " 48759 0.92 0.92 0.92 12\n",
+ " 48760 0.73 0.69 0.71 16\n",
+ " 48761 0.83 0.71 0.77 7\n",
+ " 48762 0.00 0.00 0.00 7\n",
+ " 49265 0.00 0.00 0.00 1\n",
+ " 49266 0.00 0.00 0.00 3\n",
+ " 49267 1.00 0.50 0.67 2\n",
+ " 49268 1.00 0.88 0.93 16\n",
+ " 49269 0.50 0.40 0.44 5\n",
+ " 49270 0.40 0.67 0.50 3\n",
+ " 49271 0.25 0.50 0.33 2\n",
+ " 49272 0.53 0.72 0.61 40\n",
+ " 49273 0.00 0.00 0.00 2\n",
+ " 49274 0.73 0.73 0.73 11\n",
+ " 49776 0.00 0.00 0.00 1\n",
+ " 49778 0.00 0.00 0.00 1\n",
+ " 49779 0.00 0.00 0.00 3\n",
+ " 49780 0.86 0.94 0.90 70\n",
+ " 49781 0.67 1.00 0.80 4\n",
+ " 49782 0.00 0.00 0.00 4\n",
+ " 49783 0.00 0.00 0.00 0\n",
+ " 49784 0.85 0.63 0.73 65\n",
+ " 49785 0.78 0.94 0.85 33\n",
+ " 50283 0.00 0.00 0.00 1\n",
+ " 50288 1.00 1.00 1.00 2\n",
+ " 50289 0.00 0.00 0.00 0\n",
+ " 50290 0.00 0.00 0.00 1\n",
+ " 50291 0.44 0.36 0.40 11\n",
+ " 50292 1.00 0.29 0.44 7\n",
+ " 50293 0.50 1.00 0.67 2\n",
+ " 50295 0.50 0.33 0.40 3\n",
+ " 50296 0.50 0.60 0.55 10\n",
+ " 50297 1.00 0.70 0.82 10\n",
+ " 50801 1.00 0.75 0.86 4\n",
+ " 50802 0.00 0.00 0.00 2\n",
+ " 50804 0.00 0.00 0.00 2\n",
+ " 50805 1.00 0.33 0.50 3\n",
+ " 50806 1.00 1.00 1.00 2\n",
+ " 50807 0.33 0.33 0.33 3\n",
+ " 50808 1.00 0.17 0.29 6\n",
+ " 50809 0.00 0.00 0.00 0\n",
+ " 51315 0.00 0.00 0.00 1\n",
+ " 51316 1.00 1.00 1.00 2\n",
+ " 51317 0.00 0.00 0.00 1\n",
+ " 51318 1.00 1.00 1.00 2\n",
+ " 51320 0.00 0.00 0.00 1\n",
+ " 51827 0.00 0.00 0.00 1\n",
+ " 51828 0.00 0.00 0.00 1\n",
+ " 51829 0.00 0.00 0.00 1\n",
+ " 51830 0.75 0.75 0.75 4\n",
+ " 52340 0.50 1.00 0.67 5\n",
+ " 52341 0.68 0.77 0.72 22\n",
+ " 52851 0.00 0.00 0.00 1\n",
+ " 52852 1.00 0.33 0.50 3\n",
+ " 52853 1.00 0.43 0.60 7\n",
+ " 52854 0.00 0.00 0.00 4\n",
+ " 53362 0.89 1.00 0.94 8\n",
+ " 53364 0.00 0.00 0.00 1\n",
+ " 53875 0.00 0.00 0.00 2\n",
+ " 54387 0.00 0.00 0.00 0\n",
+ " 54388 0.00 0.00 0.00 1\n",
+ " 54899 0.00 0.00 0.00 1\n",
+ " 54900 0.76 0.76 0.76 17\n",
+ " 54901 0.78 0.64 0.70 11\n",
+ " 54902 1.00 1.00 1.00 8\n",
+ " 55412 0.00 0.00 0.00 1\n",
+ " 55413 1.00 0.67 0.80 9\n",
+ "\n",
+ " accuracy 0.72 2318\n",
+ " macro avg 0.41 0.41 0.39 2318\n",
+ "weighted avg 0.70 0.72 0.70 2318\n",
+ "\n",
+ "TRAIN AND EVAL decision-tree\n",
+ "Best Parameters : {'criterion': 'gini'}\n",
+ " precision recall f1-score support\n",
+ "\n",
+ " 23545 0.00 0.00 0.00 10\n",
+ " 23546 0.50 0.71 0.59 31\n",
+ " 23547 0.00 0.00 0.00 6\n",
+ " 23981 0.00 0.00 0.00 11\n",
+ " 23982 0.89 0.88 0.88 557\n",
+ " 23983 0.65 0.66 0.65 111\n",
+ " 23984 0.00 0.00 0.00 2\n",
+ " 24422 0.65 0.73 0.69 71\n",
+ " 24423 0.52 0.49 0.50 105\n",
+ " 24424 0.31 0.33 0.32 15\n",
+ " 24866 0.00 0.00 0.00 0\n",
+ " 24867 0.71 0.73 0.72 30\n",
+ " 24868 0.14 0.17 0.15 6\n",
+ " 39537 1.00 1.00 1.00 3\n",
+ " 40045 0.00 0.00 0.00 1\n",
+ " 40047 0.00 0.00 0.00 1\n",
+ " 40558 0.67 0.50 0.57 4\n",
+ " 40559 0.40 1.00 0.57 2\n",
+ " 40560 0.00 0.00 0.00 0\n",
+ " 40561 0.00 0.00 0.00 1\n",
+ " 41070 0.00 0.00 0.00 1\n",
+ " 41071 1.00 0.75 0.86 4\n",
+ " 41072 0.00 0.00 0.00 4\n",
+ " 41073 0.00 0.00 0.00 2\n",
+ " 41582 0.00 0.00 0.00 0\n",
+ " 41583 0.83 1.00 0.90 19\n",
+ " 41584 0.00 0.00 0.00 2\n",
+ " 41585 0.00 0.00 0.00 1\n",
+ " 42095 0.20 0.33 0.25 3\n",
+ " 42096 1.00 0.50 0.67 2\n",
+ " 42097 0.00 0.00 0.00 4\n",
+ " 42606 0.00 0.00 0.00 3\n",
+ " 42607 1.00 0.50 0.67 4\n",
+ " 42608 0.00 0.00 0.00 0\n",
+ " 42609 0.33 0.33 0.33 3\n",
+ " 43117 0.00 0.00 0.00 2\n",
+ " 43118 0.33 0.50 0.40 2\n",
+ " 43119 0.33 1.00 0.50 1\n",
+ " 43120 0.00 0.00 0.00 3\n",
+ " 43121 0.50 0.33 0.40 3\n",
+ " 43122 0.00 0.00 0.00 2\n",
+ " 43630 0.00 0.00 0.00 2\n",
+ " 43631 1.00 0.25 0.40 4\n",
+ " 43632 0.00 0.00 0.00 1\n",
+ " 43633 0.00 0.00 0.00 1\n",
+ " 43634 0.00 0.00 0.00 1\n",
+ " 43635 0.00 0.00 0.00 2\n",
+ " 43636 0.80 0.36 0.50 11\n",
+ " 43637 0.00 0.00 0.00 1\n",
+ " 43638 1.00 0.50 0.67 2\n",
+ " 43639 0.50 0.33 0.40 3\n",
+ " 43640 0.00 0.00 0.00 2\n",
+ " 43641 0.00 0.00 0.00 1\n",
+ " 44142 0.67 0.67 0.67 3\n",
+ " 44143 0.70 0.70 0.70 10\n",
+ " 44144 0.00 0.00 0.00 1\n",
+ " 44146 0.00 0.00 0.00 1\n",
+ " 44147 0.00 0.00 0.00 2\n",
+ " 44148 0.00 0.00 0.00 0\n",
+ " 44149 0.00 0.00 0.00 2\n",
+ " 44150 0.53 0.50 0.52 16\n",
+ " 44151 1.00 0.57 0.73 7\n",
+ " 44152 0.57 1.00 0.73 4\n",
+ " 44153 0.00 0.00 0.00 1\n",
+ " 44154 0.00 0.00 0.00 1\n",
+ " 44654 0.33 1.00 0.50 1\n",
+ " 44655 0.75 0.75 0.75 4\n",
+ " 44656 0.00 0.00 0.00 2\n",
+ " 44657 0.00 0.00 0.00 2\n",
+ " 44658 1.00 1.00 1.00 2\n",
+ " 44660 0.40 0.50 0.44 4\n",
+ " 44661 0.80 0.44 0.57 9\n",
+ " 44662 0.65 0.70 0.68 64\n",
+ " 44663 0.37 0.58 0.45 12\n",
+ " 44664 0.50 0.33 0.40 3\n",
+ " 44665 0.33 0.33 0.33 3\n",
+ " 44666 0.40 0.67 0.50 3\n",
+ " 45165 1.00 1.00 1.00 1\n",
+ " 45167 0.00 0.00 0.00 0\n",
+ " 45168 0.00 0.00 0.00 2\n",
+ " 45169 1.00 1.00 1.00 1\n",
+ " 45170 0.33 0.50 0.40 2\n",
+ " 45171 0.00 0.00 0.00 2\n",
+ " 45172 0.57 0.44 0.50 9\n",
+ " 45173 0.68 0.70 0.69 57\n",
+ " 45174 0.78 0.82 0.80 99\n",
+ " 45175 0.29 0.29 0.29 7\n",
+ " 45176 0.17 0.33 0.22 3\n",
+ " 45177 1.00 0.17 0.29 6\n",
+ " 45678 0.00 0.00 0.00 3\n",
+ " 45679 0.67 0.29 0.40 7\n",
+ " 45680 0.00 0.00 0.00 1\n",
+ " 45681 0.00 0.00 0.00 1\n",
+ " 45682 0.62 0.62 0.62 8\n",
+ " 45683 0.00 0.00 0.00 4\n",
+ " 45684 0.00 0.00 0.00 0\n",
+ " 45685 0.36 0.36 0.36 11\n",
+ " 45686 0.46 0.48 0.47 23\n",
+ " 45687 0.50 0.42 0.45 12\n",
+ " 45688 0.75 0.43 0.55 7\n",
+ " 45689 0.72 0.81 0.76 16\n",
+ " 45690 0.83 1.00 0.91 5\n",
+ " 46184 0.25 0.33 0.29 3\n",
+ " 46189 0.00 0.00 0.00 1\n",
+ " 46190 0.75 0.75 0.75 8\n",
+ " 46191 0.00 0.00 0.00 1\n",
+ " 46192 0.00 0.00 0.00 1\n",
+ " 46193 1.00 1.00 1.00 2\n",
+ " 46194 0.33 1.00 0.50 1\n",
+ " 46195 0.00 0.00 0.00 4\n",
+ " 46196 0.00 0.00 0.00 3\n",
+ " 46197 0.00 0.00 0.00 7\n",
+ " 46198 0.25 0.20 0.22 5\n",
+ " 46199 0.11 0.20 0.14 5\n",
+ " 46200 0.47 0.56 0.51 16\n",
+ " 46201 0.80 0.44 0.57 9\n",
+ " 46697 0.69 0.69 0.69 13\n",
+ " 46701 0.00 0.00 0.00 2\n",
+ " 46702 0.00 0.00 0.00 1\n",
+ " 46703 0.00 0.00 0.00 1\n",
+ " 46704 0.00 0.00 0.00 0\n",
+ " 46705 0.00 0.00 0.00 0\n",
+ " 46707 1.00 0.50 0.67 4\n",
+ " 46709 0.33 0.17 0.22 6\n",
+ " 46710 0.42 0.36 0.38 14\n",
+ " 46711 0.43 0.50 0.46 6\n",
+ " 46712 0.00 0.00 0.00 4\n",
+ " 46713 0.00 0.00 0.00 3\n",
+ " 46714 0.33 1.00 0.50 1\n",
+ " 47212 0.67 1.00 0.80 2\n",
+ " 47213 0.00 0.00 0.00 1\n",
+ " 47214 0.00 0.00 0.00 1\n",
+ " 47215 0.00 0.00 0.00 0\n",
+ " 47217 1.00 0.25 0.40 4\n",
+ " 47218 1.00 1.00 1.00 1\n",
+ " 47219 1.00 0.50 0.67 2\n",
+ " 47220 1.00 0.50 0.67 4\n",
+ " 47221 0.40 0.62 0.48 13\n",
+ " 47222 0.50 0.17 0.25 6\n",
+ " 47223 0.50 0.50 0.50 4\n",
+ " 47224 1.00 1.00 1.00 1\n",
+ " 47225 1.00 0.56 0.71 9\n",
+ " 47226 0.00 0.00 0.00 1\n",
+ " 47723 0.00 0.00 0.00 2\n",
+ " 47726 0.00 0.00 0.00 1\n",
+ " 47729 0.00 0.00 0.00 0\n",
+ " 47730 0.00 0.00 0.00 1\n",
+ " 47731 0.00 0.00 0.00 0\n",
+ " 47732 0.00 0.00 0.00 0\n",
+ " 47733 0.56 0.42 0.48 12\n",
+ " 47734 0.00 0.00 0.00 4\n",
+ " 47735 0.33 0.50 0.40 2\n",
+ " 47736 0.00 0.00 0.00 1\n",
+ " 47737 0.00 0.00 0.00 2\n",
+ " 47738 0.00 0.00 0.00 3\n",
+ " 48235 0.50 1.00 0.67 1\n",
+ " 48236 0.00 0.00 0.00 1\n",
+ " 48237 1.00 0.50 0.67 2\n",
+ " 48239 0.00 0.00 0.00 1\n",
+ " 48240 0.00 0.00 0.00 1\n",
+ " 48243 0.00 0.00 0.00 1\n",
+ " 48244 0.38 0.50 0.43 6\n",
+ " 48245 0.40 0.20 0.27 10\n",
+ " 48246 0.50 0.20 0.29 5\n",
+ " 48247 1.00 0.50 0.67 2\n",
+ " 48248 0.29 0.67 0.40 3\n",
+ " 48249 0.50 0.50 0.50 2\n",
+ " 48250 0.00 0.00 0.00 3\n",
+ " 48748 0.00 0.00 0.00 1\n",
+ " 48749 0.00 0.00 0.00 1\n",
+ " 48753 0.00 0.00 0.00 3\n",
+ " 48754 0.00 0.00 0.00 4\n",
+ " 48755 0.00 0.00 0.00 1\n",
+ " 48756 0.17 0.67 0.27 3\n",
+ " 48757 0.93 0.94 0.93 68\n",
+ " 48758 0.00 0.00 0.00 2\n",
+ " 48759 0.71 0.83 0.77 12\n",
+ " 48760 0.83 0.31 0.45 16\n",
+ " 48761 0.20 0.14 0.17 7\n",
+ " 48762 0.50 0.29 0.36 7\n",
+ " 49259 0.00 0.00 0.00 0\n",
+ " 49265 0.00 0.00 0.00 1\n",
+ " 49266 0.00 0.00 0.00 3\n",
+ " 49267 0.00 0.00 0.00 2\n",
+ " 49268 0.81 0.81 0.81 16\n",
+ " 49269 0.00 0.00 0.00 5\n",
+ " 49270 0.00 0.00 0.00 3\n",
+ " 49271 1.00 0.50 0.67 2\n",
+ " 49272 0.54 0.62 0.58 40\n",
+ " 49273 0.07 0.50 0.12 2\n",
+ " 49274 0.71 0.45 0.56 11\n",
+ " 49776 0.00 0.00 0.00 1\n",
+ " 49778 0.00 0.00 0.00 1\n",
+ " 49779 0.00 0.00 0.00 3\n",
+ " 49780 0.89 0.89 0.89 70\n",
+ " 49781 0.60 0.75 0.67 4\n",
+ " 49782 0.00 0.00 0.00 4\n",
+ " 49783 0.00 0.00 0.00 0\n",
+ " 49784 0.68 0.63 0.66 65\n",
+ " 49785 0.60 0.55 0.57 33\n",
+ " 50283 0.00 0.00 0.00 1\n",
+ " 50288 1.00 1.00 1.00 2\n",
+ " 50289 0.00 0.00 0.00 0\n",
+ " 50290 0.00 0.00 0.00 1\n",
+ " 50291 0.44 0.36 0.40 11\n",
+ " 50292 0.80 0.57 0.67 7\n",
+ " 50293 0.00 0.00 0.00 2\n",
+ " 50295 0.50 0.33 0.40 3\n",
+ " 50296 0.27 0.30 0.29 10\n",
+ " 50297 1.00 0.80 0.89 10\n",
+ " 50801 0.00 0.00 0.00 4\n",
+ " 50802 0.00 0.00 0.00 2\n",
+ " 50803 0.00 0.00 0.00 0\n",
+ " 50804 0.50 1.00 0.67 2\n",
+ " 50805 0.67 0.67 0.67 3\n",
+ " 50806 1.00 1.00 1.00 2\n",
+ " 50807 0.50 0.67 0.57 3\n",
+ " 50808 0.33 0.17 0.22 6\n",
+ " 50809 0.00 0.00 0.00 0\n",
+ " 51314 0.00 0.00 0.00 0\n",
+ " 51315 0.00 0.00 0.00 1\n",
+ " 51316 1.00 0.50 0.67 2\n",
+ " 51317 0.00 0.00 0.00 1\n",
+ " 51318 0.00 0.00 0.00 2\n",
+ " 51320 0.00 0.00 0.00 1\n",
+ " 51826 0.00 0.00 0.00 0\n",
+ " 51827 0.00 0.00 0.00 1\n",
+ " 51828 0.00 0.00 0.00 1\n",
+ " 51829 0.00 0.00 0.00 1\n",
+ " 51830 0.50 0.25 0.33 4\n",
+ " 52339 0.00 0.00 0.00 0\n",
+ " 52340 0.67 0.80 0.73 5\n",
+ " 52341 0.56 0.68 0.61 22\n",
+ " 52851 0.00 0.00 0.00 1\n",
+ " 52852 1.00 0.33 0.50 3\n",
+ " 52853 1.00 0.57 0.73 7\n",
+ " 52854 0.00 0.00 0.00 4\n",
+ " 53362 0.73 1.00 0.84 8\n",
+ " 53363 0.00 0.00 0.00 0\n",
+ " 53364 0.00 0.00 0.00 1\n",
+ " 53875 0.00 0.00 0.00 2\n",
+ " 53876 0.00 0.00 0.00 0\n",
+ " 53877 0.00 0.00 0.00 0\n",
+ " 54387 0.00 0.00 0.00 0\n",
+ " 54388 0.00 0.00 0.00 1\n",
+ " 54389 0.00 0.00 0.00 0\n",
+ " 54899 0.00 0.00 0.00 1\n",
+ " 54900 0.65 0.65 0.65 17\n",
+ " 54901 0.57 0.36 0.44 11\n",
+ " 54902 1.00 1.00 1.00 8\n",
+ " 55412 0.11 1.00 0.20 1\n",
+ " 55413 0.78 0.78 0.78 9\n",
+ "\n",
+ " accuracy 0.63 2318\n",
+ " macro avg 0.32 0.30 0.29 2318\n",
+ "weighted avg 0.65 0.63 0.63 2318\n",
+ "\n",
+ "TRAIN AND EVAL random-forest\n"
+ ]
+ },
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/model_selection/_split.py:670: UserWarning: The least populated class in y has only 1 members, which is less than n_splits=5.\n",
+ " warnings.warn((\"The least populated class in y has only %d\"\n"
+ ]
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Best Parameters : {'criterion': 'gini', 'n_estimators': 100}\n",
+ " precision recall f1-score support\n",
+ "\n",
+ " 23545 0.33 0.10 0.15 10\n",
+ " 23546 0.68 0.68 0.68 31\n",
+ " 23547 0.00 0.00 0.00 6\n",
+ " 23981 0.00 0.00 0.00 11\n",
+ " 23982 0.82 0.97 0.89 557\n",
+ " 23983 0.79 0.64 0.71 111\n",
+ " 23984 0.00 0.00 0.00 2\n",
+ " 24422 0.82 0.75 0.78 71\n",
+ " 24423 0.65 0.64 0.64 105\n",
+ " 24424 0.40 0.13 0.20 15\n",
+ " 24867 0.80 0.80 0.80 30\n",
+ " 24868 1.00 0.33 0.50 6\n",
+ " 39537 1.00 0.67 0.80 3\n",
+ " 40045 0.00 0.00 0.00 1\n",
+ " 40046 0.00 0.00 0.00 0\n",
+ " 40047 0.00 0.00 0.00 1\n",
+ " 40049 0.00 0.00 0.00 0\n",
+ " 40558 1.00 0.50 0.67 4\n",
+ " 40559 0.67 1.00 0.80 2\n",
+ " 40561 0.00 0.00 0.00 1\n",
+ " 41070 0.00 0.00 0.00 1\n",
+ " 41071 1.00 0.75 0.86 4\n",
+ " 41072 0.00 0.00 0.00 4\n",
+ " 41073 0.25 0.50 0.33 2\n",
+ " 41582 0.00 0.00 0.00 0\n",
+ " 41583 0.90 1.00 0.95 19\n",
+ " 41584 0.33 0.50 0.40 2\n",
+ " 41585 0.00 0.00 0.00 1\n",
+ " 42094 0.00 0.00 0.00 0\n",
+ " 42095 0.25 0.33 0.29 3\n",
+ " 42096 1.00 1.00 1.00 2\n",
+ " 42097 0.00 0.00 0.00 4\n",
+ " 42606 0.00 0.00 0.00 3\n",
+ " 42607 1.00 0.25 0.40 4\n",
+ " 42609 0.50 0.33 0.40 3\n",
+ " 43117 0.00 0.00 0.00 2\n",
+ " 43118 0.50 0.50 0.50 2\n",
+ " 43119 0.00 0.00 0.00 1\n",
+ " 43120 0.00 0.00 0.00 3\n",
+ " 43121 1.00 0.67 0.80 3\n",
+ " 43122 0.33 0.50 0.40 2\n",
+ " 43123 0.00 0.00 0.00 0\n",
+ " 43630 0.00 0.00 0.00 2\n",
+ " 43631 0.20 0.25 0.22 4\n",
+ " 43632 0.00 0.00 0.00 1\n",
+ " 43633 0.00 0.00 0.00 1\n",
+ " 43634 0.00 0.00 0.00 1\n",
+ " 43635 0.00 0.00 0.00 2\n",
+ " 43636 0.78 0.64 0.70 11\n",
+ " 43637 0.00 0.00 0.00 1\n",
+ " 43638 1.00 0.50 0.67 2\n",
+ " 43639 0.00 0.00 0.00 3\n",
+ " 43640 0.00 0.00 0.00 2\n",
+ " 43641 0.00 0.00 0.00 1\n",
+ " 44142 0.40 0.67 0.50 3\n",
+ " 44143 0.73 0.80 0.76 10\n",
+ " 44144 0.50 1.00 0.67 1\n",
+ " 44146 0.00 0.00 0.00 1\n",
+ " 44147 0.00 0.00 0.00 2\n",
+ " 44149 0.00 0.00 0.00 2\n",
+ " 44150 0.68 0.81 0.74 16\n",
+ " 44151 0.57 0.57 0.57 7\n",
+ " 44152 0.57 1.00 0.73 4\n",
+ " 44153 0.00 0.00 0.00 1\n",
+ " 44154 0.00 0.00 0.00 1\n",
+ " 44654 0.00 0.00 0.00 1\n",
+ " 44655 0.75 0.75 0.75 4\n",
+ " 44656 0.00 0.00 0.00 2\n",
+ " 44657 0.00 0.00 0.00 2\n",
+ " 44658 1.00 1.00 1.00 2\n",
+ " 44660 0.00 0.00 0.00 4\n",
+ " 44661 0.83 0.56 0.67 9\n",
+ " 44662 0.81 0.78 0.79 64\n",
+ " 44663 0.33 0.50 0.40 12\n",
+ " 44664 0.50 0.33 0.40 3\n",
+ " 44665 0.33 0.33 0.33 3\n",
+ " 44666 0.40 0.67 0.50 3\n",
+ " 45165 1.00 1.00 1.00 1\n",
+ " 45168 1.00 0.50 0.67 2\n",
+ " 45169 0.50 1.00 0.67 1\n",
+ " 45170 0.50 0.50 0.50 2\n",
+ " 45171 0.00 0.00 0.00 2\n",
+ " 45172 0.60 0.33 0.43 9\n",
+ " 45173 0.78 0.86 0.82 57\n",
+ " 45174 0.67 0.84 0.75 99\n",
+ " 45175 0.33 0.14 0.20 7\n",
+ " 45176 0.50 1.00 0.67 3\n",
+ " 45177 0.50 0.17 0.25 6\n",
+ " 45678 0.00 0.00 0.00 3\n",
+ " 45679 1.00 0.29 0.44 7\n",
+ " 45680 0.00 0.00 0.00 1\n",
+ " 45681 0.00 0.00 0.00 1\n",
+ " 45682 0.78 0.88 0.82 8\n",
+ " 45683 0.00 0.00 0.00 4\n",
+ " 45684 0.00 0.00 0.00 0\n",
+ " 45685 0.70 0.64 0.67 11\n",
+ " 45686 0.50 0.48 0.49 23\n",
+ " 45687 0.64 0.75 0.69 12\n",
+ " 45688 0.33 0.29 0.31 7\n",
+ " 45689 0.62 0.81 0.70 16\n",
+ " 45690 0.80 0.80 0.80 5\n",
+ " 46184 0.00 0.00 0.00 3\n",
+ " 46189 0.00 0.00 0.00 1\n",
+ " 46190 0.56 0.62 0.59 8\n",
+ " 46191 0.00 0.00 0.00 1\n",
+ " 46192 0.00 0.00 0.00 1\n",
+ " 46193 1.00 1.00 1.00 2\n",
+ " 46194 0.00 0.00 0.00 1\n",
+ " 46195 0.00 0.00 0.00 4\n",
+ " 46196 1.00 0.33 0.50 3\n",
+ " 46197 0.17 0.14 0.15 7\n",
+ " 46198 0.60 0.60 0.60 5\n",
+ " 46199 0.50 0.40 0.44 5\n",
+ " 46200 0.57 0.75 0.65 16\n",
+ " 46201 1.00 0.33 0.50 9\n",
+ " 46697 0.80 0.92 0.86 13\n",
+ " 46701 1.00 1.00 1.00 2\n",
+ " 46702 0.00 0.00 0.00 1\n",
+ " 46703 0.00 0.00 0.00 1\n",
+ " 46707 0.67 0.50 0.57 4\n",
+ " 46709 0.33 0.17 0.22 6\n",
+ " 46710 0.42 0.36 0.38 14\n",
+ " 46711 0.60 0.50 0.55 6\n",
+ " 46712 0.00 0.00 0.00 4\n",
+ " 46713 0.00 0.00 0.00 3\n",
+ " 46714 0.50 1.00 0.67 1\n",
+ " 47210 0.00 0.00 0.00 0\n",
+ " 47211 0.00 0.00 0.00 0\n",
+ " 47212 0.33 0.50 0.40 2\n",
+ " 47213 0.50 1.00 0.67 1\n",
+ " 47214 0.00 0.00 0.00 1\n",
+ " 47217 1.00 0.50 0.67 4\n",
+ " 47218 0.00 0.00 0.00 1\n",
+ " 47219 1.00 0.50 0.67 2\n",
+ " 47220 0.75 0.75 0.75 4\n",
+ " 47221 0.40 0.62 0.48 13\n",
+ " 47222 0.50 0.33 0.40 6\n",
+ " 47223 0.50 0.50 0.50 4\n",
+ " 47224 0.50 1.00 0.67 1\n",
+ " 47225 0.89 0.89 0.89 9\n",
+ " 47226 0.50 1.00 0.67 1\n",
+ " 47723 1.00 0.50 0.67 2\n",
+ " 47726 0.00 0.00 0.00 1\n",
+ " 47730 0.00 0.00 0.00 1\n",
+ " 47731 0.00 0.00 0.00 0\n",
+ " 47733 0.55 0.50 0.52 12\n",
+ " 47734 0.25 0.25 0.25 4\n",
+ " 47735 0.25 0.50 0.33 2\n",
+ " 47736 0.00 0.00 0.00 1\n",
+ " 47737 0.00 0.00 0.00 2\n",
+ " 47738 0.20 0.33 0.25 3\n",
+ " 48235 0.00 0.00 0.00 1\n",
+ " 48236 0.00 0.00 0.00 1\n",
+ " 48237 1.00 1.00 1.00 2\n",
+ " 48239 0.00 0.00 0.00 1\n",
+ " 48240 0.00 0.00 0.00 1\n",
+ " 48242 0.00 0.00 0.00 0\n",
+ " 48243 0.00 0.00 0.00 1\n",
+ " 48244 1.00 0.50 0.67 6\n",
+ " 48245 0.40 0.20 0.27 10\n",
+ " 48246 0.00 0.00 0.00 5\n",
+ " 48247 1.00 0.50 0.67 2\n",
+ " 48248 0.67 0.67 0.67 3\n",
+ " 48249 0.25 0.50 0.33 2\n",
+ " 48250 0.00 0.00 0.00 3\n",
+ " 48748 0.00 0.00 0.00 1\n",
+ " 48749 0.00 0.00 0.00 1\n",
+ " 48753 0.00 0.00 0.00 3\n",
+ " 48754 0.00 0.00 0.00 4\n",
+ " 48755 0.00 0.00 0.00 1\n",
+ " 48756 0.40 0.67 0.50 3\n",
+ " 48757 0.85 0.99 0.91 68\n",
+ " 48758 0.00 0.00 0.00 2\n",
+ " 48759 0.71 0.83 0.77 12\n",
+ " 48760 0.50 0.31 0.38 16\n",
+ " 48761 0.50 0.29 0.36 7\n",
+ " 48762 1.00 0.43 0.60 7\n",
+ " 49265 0.00 0.00 0.00 1\n",
+ " 49266 0.00 0.00 0.00 3\n",
+ " 49267 0.00 0.00 0.00 2\n",
+ " 49268 0.88 0.94 0.91 16\n",
+ " 49269 0.00 0.00 0.00 5\n",
+ " 49270 0.40 0.67 0.50 3\n",
+ " 49271 0.25 0.50 0.33 2\n",
+ " 49272 0.59 0.68 0.63 40\n",
+ " 49273 0.17 0.50 0.25 2\n",
+ " 49274 0.80 0.73 0.76 11\n",
+ " 49773 0.00 0.00 0.00 0\n",
+ " 49776 0.00 0.00 0.00 1\n",
+ " 49778 0.00 0.00 0.00 1\n",
+ " 49779 1.00 0.33 0.50 3\n",
+ " 49780 0.85 0.99 0.91 70\n",
+ " 49781 0.60 0.75 0.67 4\n",
+ " 49782 0.00 0.00 0.00 4\n",
+ " 49783 0.00 0.00 0.00 0\n",
+ " 49784 0.79 0.74 0.76 65\n",
+ " 49785 0.64 0.88 0.74 33\n",
+ " 50283 0.00 0.00 0.00 1\n",
+ " 50288 1.00 1.00 1.00 2\n",
+ " 50289 0.00 0.00 0.00 0\n",
+ " 50290 0.00 0.00 0.00 1\n",
+ " 50291 0.50 0.27 0.35 11\n",
+ " 50292 0.67 0.29 0.40 7\n",
+ " 50293 0.33 0.50 0.40 2\n",
+ " 50294 0.00 0.00 0.00 0\n",
+ " 50295 0.50 0.33 0.40 3\n",
+ " 50296 0.27 0.30 0.29 10\n",
+ " 50297 1.00 0.70 0.82 10\n",
+ " 50801 1.00 0.50 0.67 4\n",
+ " 50802 0.00 0.00 0.00 2\n",
+ " 50804 0.00 0.00 0.00 2\n",
+ " 50805 0.50 0.67 0.57 3\n",
+ " 50806 0.67 1.00 0.80 2\n",
+ " 50807 0.50 0.33 0.40 3\n",
+ " 50808 0.20 0.17 0.18 6\n",
+ " 50809 0.00 0.00 0.00 0\n",
+ " 51315 0.00 0.00 0.00 1\n",
+ " 51316 1.00 1.00 1.00 2\n",
+ " 51317 0.00 0.00 0.00 1\n",
+ " 51318 1.00 1.00 1.00 2\n",
+ " 51319 0.00 0.00 0.00 0\n",
+ " 51320 0.00 0.00 0.00 1\n",
+ " 51827 0.00 0.00 0.00 1\n",
+ " 51828 0.00 0.00 0.00 1\n",
+ " 51829 0.00 0.00 0.00 1\n",
+ " 51830 0.50 0.25 0.33 4\n",
+ " 52340 0.50 0.80 0.62 5\n",
+ " 52341 0.65 0.68 0.67 22\n",
+ " 52851 0.00 0.00 0.00 1\n",
+ " 52852 1.00 0.33 0.50 3\n",
+ " 52853 0.80 0.57 0.67 7\n",
+ " 52854 1.00 0.50 0.67 4\n",
+ " 53362 0.80 1.00 0.89 8\n",
+ " 53364 0.00 0.00 0.00 1\n",
+ " 53875 0.00 0.00 0.00 2\n",
+ " 54387 0.00 0.00 0.00 0\n",
+ " 54388 0.00 0.00 0.00 1\n",
+ " 54389 0.00 0.00 0.00 0\n",
+ " 54899 0.00 0.00 0.00 1\n",
+ " 54900 0.76 0.76 0.76 17\n",
+ " 54901 0.70 0.64 0.67 11\n",
+ " 54902 1.00 0.88 0.93 8\n",
+ " 55412 0.20 1.00 0.33 1\n",
+ " 55413 0.67 0.67 0.67 9\n",
+ "\n",
+ " accuracy 0.70 2318\n",
+ " macro avg 0.37 0.35 0.34 2318\n",
+ "weighted avg 0.68 0.70 0.68 2318\n",
+ "\n"
+ ]
+ },
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n",
+ "/usr/local/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. Use `zero_division` parameter to control this behavior.\n",
+ " _warn_prf(average, modifier, msg_start, len(result))\n"
+ ]
+ }
+ ],
+ "source": [
+ "for CLASSIFIER in classifier_dict:\n",
+ " print(\"TRAIN AND EVAL {0}\".format(CLASSIFIER))\n",
+ " clf = GridSearchCV(\n",
+ " classifier_dict[CLASSIFIER], parameters[CLASSIFIER], scoring='f1_weighted',n_jobs=-1\n",
+ " )\n",
+ " clf.fit(X_train, y_train)\n",
+ " print(\"Best Parameters : \",clf.best_params_)\n",
+ " y_pred = clf.best_estimator_.predict(X_test)\n",
+ " print(classification_report(y_test,y_pred))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3.7.5 64-bit",
+ "language": "python",
+ "name": "python37564bitdc8b0e1290e74b85b0e630c435ea2fe8"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.5"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/notebooks/disambiguate.ipynb b/notebooks/disambiguate.ipynb
new file mode 100644
index 0000000..7bf82f4
--- /dev/null
+++ b/notebooks/disambiguate.ipynb
@@ -0,0 +1,143 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import glob\n",
+ "import pandas as pd\n",
+ "import numpy as np\n",
+ "from tqdm.notebook import tqdm"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "'\\ng = Geocoder(\"outputs/FR_MODEL_2/FR.txt_100_4_100__A_C.h5\",\"outputs/FR_MODEL_2/FR.txt_100_4_100__A_C_index\")\\nprint(g.get_coord(\"Paris\",\"France\"))\\ng.get_coords?? #lon lat\\n'"
+ ]
+ },
+ "execution_count": 2,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "from lib.geocoder.our_geocoder import Geocoder\n",
+ "from lib.utils_geo import haversine_pd\n",
+ "\"\"\"\n",
+ "g = Geocoder(\"outputs/FR_MODEL_2/FR.txt_100_4_100__A_C.h5\",\"outputs/FR_MODEL_2/FR.txt_100_4_100__A_C_index\")\n",
+ "print(g.get_coord(\"Paris\",\"France\"))\n",
+ "g.get_coords?? #lon lat\n",
+ "\"\"\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "basedir = \"outputs/FR_MODEL_2/\"\n",
+ "geocoding_df = pd.read_csv(\"geocoding_data/ambig_data/FR.csv\",sep=\"\\t\",index_col=0)\n",
+ "geocoding_df = geocoding_df[geocoding_df.split == \"test\"]"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "['FR.txt_100_4_100__A_I_C',\n",
+ " 'FR.txt_100_4_100__A_C',\n",
+ " 'FR.txt_100_4_100__C',\n",
+ " 'FR.txt_100_4_100__I_C']"
+ ]
+ },
+ "execution_count": 8,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "model_available = glob.glob(basedir+\"*.h5\")\n",
+ "model_available = [mod.rstrip(\".h5\").split(\"/\")[-1] for mod in model_available]\n",
+ "model_available"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 9,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "def accuracy_at_k(geocoding_df,geocoder,k=100):\n",
+ " lons,lats = g.get_coords(geocoding_df.toponym.values,geocoding_df.toponym_context.values)\n",
+ " geocoding_df[\"pred_latitude\"] = lats\n",
+ " geocoding_df[\"pred_longitude\"] = lons\n",
+ " geocoding_df[\"distanceKM\"] = haversine_pd(geocoding_df.longitude,geocoding_df.latitude,geocoding_df.pred_longitude,geocoding_df.pred_latitude)\n",
+ " return (geocoding_df.distanceKM <k).sum()/len(geocoding_df)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "res_ = []\n",
+ "for mod in tqdm(model_available):\n",
+ " index_fn = basedir + mod +\"_index\"\n",
+ " model_fn = basedir + mod +\".h5\"\n",
+ " g = Geocoder(model_fn, index_fn)\n",
+ " res_.append([mod,accuracy_at_k(geocoding_df,g,100),accuracy_at_k(geocoding_df,g,50),accuracy_at_k(geocoding_df,g,20)])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "pd.DataFrame(res_,columns=\"dataset accuracy@100km accuracy@50km accuracy@20km\".split()).to_csv(\"US_COOC_test_result_geocoding.csv\",sep=\"\\t\",index=None)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3.7.5 64-bit",
+ "language": "python",
+ "name": "python37564bitdc8b0e1290e74b85b0e630c435ea2fe8"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.5"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/notebooks/generate_ambiguous_dataset.ipynb b/notebooks/generate_ambiguous_dataset.ipynb
new file mode 100644
index 0000000..1fb62b9
--- /dev/null
+++ b/notebooks/generate_ambiguous_dataset.ipynb
@@ -0,0 +1,273 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "\u001b[31mBO.txt\u001b[m\u001b[m \u001b[31mFR.txt_train.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt\u001b[m\u001b[m \u001b[1m\u001b[36mGB\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_11adjacency.json\u001b[m\u001b[m \u001b[31mGB.zip\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_44adjacency.json\u001b[m\u001b[m \u001b[31mME.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_52adjacency.json\u001b[m\u001b[m \u001b[31mTH.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_75adjacency.json\u001b[m\u001b[m \u001b[1m\u001b[36mUS\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_76adjacency.json\u001b[m\u001b[m \u001b[31mUS.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_84adjacency.json\u001b[m\u001b[m \u001b[31mUS.txt_test.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1_93adjacency.json\u001b[m\u001b[m \u001b[31mUS.txt_train.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_1adjacency.json\u001b[m\u001b[m \u001b[31mUS.zip\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_11adjacency.json\u001b[m\u001b[m \u001b[31mUS_FR.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_24adjacency.json\u001b[m\u001b[m \u001b[31mUS_FR.txt_1adjacency.json\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_27adjacency.json\u001b[m\u001b[m \u001b[31mUS_FR.txt_test.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_28adjacency.json\u001b[m\u001b[m \u001b[31mUS_FR.txt_train.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_32adjacency.json\u001b[m\u001b[m \u001b[31mUY.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_44adjacency.json\u001b[m\u001b[m \u001b[31mallCountries.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_52adjacency.json\u001b[m\u001b[m \u001b[31mallCountries.txt_test.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_53adjacency.json\u001b[m\u001b[m \u001b[31mallCountries.txt_train.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_75adjacency.json\u001b[m\u001b[m \u001b[31mall_countries_test.png\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_76adjacency.json\u001b[m\u001b[m \u001b[31mall_countries_train.png\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_84adjacency.json\u001b[m\u001b[m \u001b[31malternateNames.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_93adjacency.json\u001b[m\u001b[m \u001b[31mhierarchy.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_50_94adjacency.json\u001b[m\u001b[m \u001b[31mhierarchy.txt_test.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_5_Noneadjacency.json\u001b[m\u001b[m \u001b[31mhierarchy.txt_train.csv\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_5_Noneadjacency_backup.json\u001b[m\u001b[m \u001b[31miso-languagecodes.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_teeeest.csv\u001b[m\u001b[m \u001b[31mplotGeonamesDatasets.ipynb\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_test.csv\u001b[m\u001b[m \u001b[31mreadme.txt\u001b[m\u001b[m\n",
+ "\u001b[31mFR.txt_traieeen.csv\u001b[m\u001b[m \u001b[31mtest.pdf\u001b[m\u001b[m\n"
+ ]
+ }
+ ],
+ "source": []
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 49,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import pandas as pd\n",
+ "from helpers import read_geonames\n",
+ "from lib.utils_geo import latlon2healpix\n",
+ "from tqdm.notebook import tqdm"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "basedir = \"/Volumes/My Passport/SAVE_avant_confinement_2/data/geonamesData/\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 64,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "df = read_geonames(basedir + \"GB/GB.txt\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 65,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "df[\"hp_128\"] = df.apply(lambda x:latlon2healpix(x.latitude,x.longitude,128),axis=1)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 66,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "counts = df.name.value_counts().reset_index()\n",
+ "counts = counts[counts.name >5]\n",
+ "ambiguous_toponym = counts[\"index\"].values"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 67,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "5653\n"
+ ]
+ }
+ ],
+ "source": [
+ "for row in df.itertuples():\n",
+ " print(row.geonameid)\n",
+ " break"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 68,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "206"
+ ]
+ },
+ "execution_count": 68,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "len(ambiguous_toponym)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 69,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "7167aacc08a04971a07a46c61de23b2a",
+ "version_major": 2,
+ "version_minor": 0
+ },
+ "text/plain": [
+ "HBox(children=(FloatProgress(value=0.0, max=206.0), HTML(value='')))"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "\n"
+ ]
+ }
+ ],
+ "source": [
+ "pairs_of_toponym = []\n",
+ "for toponym in tqdm(ambiguous_toponym):\n",
+ " sample_ambiguous = df[df.name == toponym].sample(5)\n",
+ " for ix, row in sample_ambiguous.iterrows():\n",
+ " in_radius = df[df.hp_128 == row.hp_128].sample(1)\n",
+ " pairs_of_toponym.extend([[toponym,val,row.latitude,row.longitude,row.hp_128 ]for val in in_radius[\"name\"]])\n",
+ " "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 70,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "out_DF = pd.DataFrame(pairs_of_toponym,columns=\"toponym toponym_context latitude longitude hp_split\".split())"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 71,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "out_DF[\"split\"] = \"test\"\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 72,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "out_DF.to_csv(\"AMB_GB.csv\",sep=\"\\t\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 74,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "\ttoponym\ttoponym_context\tlatitude\tlongitude\thp_split\tsplit\n",
+ "0\tSutton\tWentworth\t52.388040000000004\t0.11865999999999999\t20200\ttest\n",
+ "1\tSutton\tPreston\t51.193540000000006\t1.32185\t21425\ttest\n",
+ "2\tSutton\tThornhaugh\t52.57666\t-0.38199\t20603\ttest\n",
+ "3\tSutton\tFelbrigg\t52.758869999999995\t1.52591\t19801\ttest\n",
+ "4\tSutton\tCretingham\t52.06667\t1.35\t20605\ttest\n",
+ "5\tMiddleton\tMiddleton\t55.63041\t-1.82285\t17110\ttest\n",
+ "6\tMiddleton\tColwick\t52.574130000000004\t-1.74308\t20198\ttest\n",
+ "7\tMiddleton\tBubwith\t53.92598\t-0.58099\t19011\ttest\n",
+ "8\tMiddleton\tCastle Howard\t54.25\t-0.8\t18623\ttest\n"
+ ]
+ }
+ ],
+ "source": [
+ "!head AMB_GB.csv"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 75,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "'/Users/jacquesfize/POSTDOCLYON'"
+ ]
+ },
+ "execution_count": 75,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "pwd"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3.7.5 64-bit",
+ "language": "python",
+ "name": "python37564bitdc8b0e1290e74b85b0e630c435ea2fe8"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.5"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/notebooks/heuristic.ipynb b/notebooks/heuristic.ipynb
new file mode 100644
index 0000000..1c83a19
--- /dev/null
+++ b/notebooks/heuristic.ipynb
@@ -0,0 +1,407 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from lib.geocoder.our_geocoder import Geocoder\n",
+ "import matplotlib.pyplot as plt\n",
+ "import geopandas as gpd"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "g = Geocoder(\"outputs/FR_MODEL_2/nside_512/FR_nside_512/FR_nside_512_4_100_A_I_P.h5.part\",\"outputs/FR_MODEL_2/nside_512/FR_nside_512/FR_nside_512_4_100_A_I_P_index\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "toponyms = \"Paris;Cherbourg;Saint-Lô;Caen;Lyon;Rennes;Montpellier;Occitanie;Toulouse;Pau\".split(\";\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from lib.geocoder.heuristics import *"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "{'Paris': {'lat': 48.891907, 'lon': 2.3173065},\n",
+ " 'Cherbourg': {'lat': 49.09114, 'lon': -1.1740417},\n",
+ " 'Saint-Lô': {'lat': 48.602127, 'lon': -0.73223877},\n",
+ " 'Caen': {'lat': 48.751816, 'lon': 0.055923462},\n",
+ " 'Lyon': {'lat': 45.72853, 'lon': 4.729294},\n",
+ " 'Rennes': {'lat': 48.395355, 'lon': -1.3266144},\n",
+ " 'Montpellier': {'lat': 43.52542, 'lon': 2.7322083},\n",
+ " 'Occitanie': {'lat': 43.17047, 'lon': 1.231308},\n",
+ " 'Toulouse': {'lat': 43.70291, 'lon': 1.5109406},\n",
+ " 'Pau': {'lat': 44.439102, 'lon': 1.7921295}}"
+ ]
+ },
+ "execution_count": 7,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "res_heuristic = heuristic_no_context(g,toponyms)\n",
+ "res_heuristic"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "(40.0, 55.0)"
+ ]
+ },
+ "execution_count": 8,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1080x1080 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "from shapely.geometry import Point\n",
+ "df_point = pd.DataFrame.from_dict(res_heuristic,orient=\"index\")\n",
+ "df_point[\"geometry\"] = df_point.apply(lambda x:Point(x.lon,x.lat),axis=1)\n",
+ "fig, ax = plt.subplots(1,figsize=(15,15))\n",
+ "world = gpd.read_file(gpd.datasets.get_path('naturalearth_lowres'))\n",
+ "world.plot(color='white', edgecolor='black',ax=ax)\n",
+ "gpd.GeoDataFrame(df_point).plot(ax=ax)\n",
+ "for tp,coord in res_heuristic.items(): \n",
+ " ax.annotate(tp,(coord[\"lon\"],coord[\"lat\"]))\n",
+ "ax.set_xlim((-10,10))\n",
+ "ax.set_ylim((40,55))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 9,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div style=\"width:100%;\"><div style=\"position:relative;width:100%;height:0;padding-bottom:60%;\"><span style=\"color:#565656\">Make this Notebook Trusted to load map: File -> Trust Notebook</span><iframe src=\"about:blank\" style=\"position:absolute;width:100%;height:100%;left:0;top:0;border:none !important;\" data-html=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 onload=\"this.contentDocument.open();this.contentDocument.write(atob(this.getAttribute('data-html')));this.contentDocument.close();\" allowfullscreen webkitallowfullscreen mozallowfullscreen></iframe></div></div>"
+ ],
+ "text/plain": [
+ "<folium.folium.Map at 0x163566b50>"
+ ]
+ },
+ "execution_count": 9,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "input_ = np.asarray([[t1,t2] for t2 in toponyms for t1 in toponyms if t2 != t1])\n",
+ "res_geocode = pd.DataFrame(input_,columns=\"t tc\".split())\n",
+ "lons,lats = g.get_coords(input_[:,0],input_[:,1])\n",
+ "res_geocode[\"lon\"] = lons\n",
+ "res_geocode[\"lat\"] = lats\n",
+ "\n",
+ "\n",
+ "from shapely.geometry import Point\n",
+ "import folium\n",
+ "m = folium.Map()\n",
+ "for ix,row in res_geocode.iterrows(): \n",
+ " ax.annotate(row.t + \" \" + row.tc,(row[\"lon\"],row[\"lat\"]))\n",
+ " folium.Marker([row.lat, row.lon], popup=row.t + \" \" + row.tc).add_to(m)\n",
+ " \n",
+ "m"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 10,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "\u001b[0;31mSignature:\u001b[0m \u001b[0mg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot_coord\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtoponym\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlat\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlon\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minteractive_map\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+ "\u001b[0;31mDocstring:\u001b[0m <no docstring>\n",
+ "\u001b[0;31mSource:\u001b[0m \n",
+ " \u001b[0;32mdef\u001b[0m \u001b[0mplot_coord\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mtoponym\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mlat\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mlon\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0minteractive_map\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0minteractive_map\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mfolium\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mtempfile\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mwebbrowser\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mfp\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtempfile\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mNamedTemporaryFile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdelete\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mm\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfolium\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mMap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mfolium\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mMarker\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mlat\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlon\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpopup\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtoponym\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madd_to\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mm\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mm\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mwebbrowser\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'file://'\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mfp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmatplotlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpyplot\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mgeopandas\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mfig\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0max\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubplots\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mworld\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgeopandas\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread_file\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgeopandas\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdatasets\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_path\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'naturalearth_lowres'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mworld\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcolor\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'white'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0medgecolor\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'black'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0max\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlon\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mlat\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mmarker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'o'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcolor\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'red'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmarkersize\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\n",
+ "\u001b[0;34m\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+ "\u001b[0;31mFile:\u001b[0m ~/POSTDOCLYON/toponym-geocoding/lib/geocoder/our_geocoder.py\n",
+ "\u001b[0;31mType:\u001b[0m method\n"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "g.plot_coord??"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 11,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "\n",
+ "\n",
+ "from lib.utils_geo import haversine_pd\n",
+ "from sklearn.cluster import DBSCAN\n",
+ "from sklearn.metrics.pairwise import haversine_distances"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 12,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "(90, 2)"
+ ]
+ },
+ "execution_count": 12,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "res_geocode[\"lon lat\".split()].values.shape"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 13,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from haversine import haversine_vector, Unit"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "array([[ 0. , 74.38099 , 31.865028, ..., 670.45294 , 644.2007 ,\n",
+ " 613.9308 ],\n",
+ " [ 74.38099 , 0. , 102.8423 , ..., 625.586 , 622.1915 ,\n",
+ " 581.8792 ],\n",
+ " [ 31.865028, 102.8423 , 0. , ..., 675.55005 , 638.657 ,\n",
+ " 613.02527 ],\n",
+ " ...,\n",
+ " [670.45294 , 625.586 , 675.55005 , ..., 0. , 221.29712 ,\n",
+ " 134.94882 ],\n",
+ " [644.2007 , 622.1915 , 638.657 , ..., 221.29712 , 0. ,\n",
+ " 95.80514 ],\n",
+ " [613.9308 , 581.8792 , 613.02527 , ..., 134.94882 , 95.80514 ,\n",
+ " 0. ]], dtype=float32)"
+ ]
+ },
+ "execution_count": 14,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "haversine_vector(res_geocode[\"lon lat\".split()].values,res_geocode[\"lon lat\".split()].values,unit=\"km\",comb=True)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 15,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "{'Paris': {'lat': 48.510147, 'lon': 0.8369751},\n",
+ " 'Cherbourg': {'lat': 47.15422, 'lon': -0.20059204},\n",
+ " 'Saint-Lô': {'lat': 47.375473, 'lon': 0.21366882},\n",
+ " 'Caen': {'lat': 48.834396, 'lon': 0.093063354},\n",
+ " 'Lyon': {'lat': 45.77159, 'lon': 3.4461823},\n",
+ " 'Rennes': {'lat': 46.499756, 'lon': -0.51908875},\n",
+ " 'Montpellier': {'lat': 43.609375, 'lon': 1.9396667},\n",
+ " 'Occitanie': {'lat': 43.59912, 'lon': 0.7354431},\n",
+ " 'Toulouse': {'lat': 43.70291, 'lon': 1.5823364},\n",
+ " 'Pau': {'lat': 46.071503, 'lon': 0.79255676}}"
+ ]
+ },
+ "execution_count": 15,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "\n",
+ "from haversine import haversine_vector, Unit\n",
+ "def heuristic_cluster(geocoder,toponyms,eps=100):\n",
+ " results = {}\n",
+ " input_ = np.asarray([[t1,t2] for t2 in toponyms for t1 in toponyms if t2 != t1])\n",
+ " res_geocode = pd.DataFrame(input_,columns=\"t tc\".split())\n",
+ " lons,lats = geocoder.get_coords(input_[:,0],input_[:,1])\n",
+ " res_geocode[\"lon\"] = lons\n",
+ " res_geocode[\"lat\"] = lats\n",
+ "\n",
+ " clf = DBSCAN(eps=eps)\n",
+ " for t in toponyms:\n",
+ " tp_df = res_geocode[res_geocode.tc == t].copy()\n",
+ "\n",
+ " coords = tp_df[\"lon lat\".split()].values\n",
+ " clf.fit(haversine_vector(coords,coords,unit=\"km\",comb=True))\n",
+ "\n",
+ " tp_df[\"cluster\"] = clf.labels_\n",
+ " counts_ = dict(tp_df.cluster.value_counts())\n",
+ " max_cluster = max(counts_, key=counts_.get)\n",
+ " tp_df = tp_df[tp_df.cluster == max_cluster]\n",
+ " lat = tp_df.lat.median()\n",
+ " lon = tp_df.lon.median() #\n",
+ " results[t]={\"lat\":lat,\"lon\":lon}\n",
+ " return results\n",
+ "\n",
+ "results = heuristic_cluster(g,toponyms)\n",
+ " \n",
+ "results\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 16,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "(30.0, 55.0)"
+ ]
+ },
+ "execution_count": 16,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1080x1080 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "from shapely.geometry import Point\n",
+ "df_point = pd.DataFrame.from_dict(results,orient=\"index\")\n",
+ "df_point[\"geometry\"] = df_point.apply(lambda x:Point(x.lon,x.lat),axis=1)\n",
+ "fig, ax = plt.subplots(1,figsize=(15,15))\n",
+ "world = gpd.read_file(gpd.datasets.get_path('naturalearth_lowres'))\n",
+ "world.plot(color='white', edgecolor='black',ax=ax)\n",
+ "gpd.GeoDataFrame(df_point).plot(ax=ax)\n",
+ "for tp,coord in results.items(): \n",
+ " ax.annotate(tp,(coord[\"lon\"],coord[\"lat\"]))\n",
+ "ax.set_xlim((-10,10))\n",
+ "ax.set_ylim((30,55))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3.7.5 64-bit",
+ "language": "python",
+ "name": "python37564bitdc8b0e1290e74b85b0e630c435ea2fe8"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.5"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/notebooks/tobert.ipynb b/notebooks/tobert.ipynb
new file mode 100644
index 0000000..18fde80
--- /dev/null
+++ b/notebooks/tobert.ipynb
@@ -0,0 +1,297 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import pandas as pd\n",
+ "from lib.utils_geo import latlon2healpix"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "basedir = \"/Volumes/My Passport/SAVE_avant_confinement_2/toponym-geocoding/data_new_/\"\n",
+ "dataset = \"TX_IDF\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 23,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "def parsetoBert(df):\n",
+ " return pd.DataFrame.from_records(df.apply(lambda x: (x.toponym + \" \" + x.toponym_context,x.hp_split,x.split) ,axis=1).values,columns=\"sentence label split\".split())\n",
+ "\n",
+ "def read_and_parse_data(basedir,dataset,type_):\n",
+ " df = pd.read_csv(\"{0}/{1}_{2}.csv\".format(basedir,dataset,type_),sep=\"\\t\",index_col=0)\n",
+ " df[\"hp_split\"] = df.apply(lambda x: latlon2healpix(x.latitude,x.longitude,128),axis=1)\n",
+ " return parsetoBert(df)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 28,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "type_list = [\n",
+ " \"inclusion\",\n",
+ " \"adjacent\",\n",
+ " \"cooc\"\n",
+ "]\n",
+ "list_combination = [\n",
+ " [2],\n",
+ " [0,1,2],\n",
+ " [0,2],\n",
+ " [1,2]\n",
+ "]\n",
+ "dataset = \"TX_IDF\"\n",
+ "for comb in list_combination:\n",
+ " res = []\n",
+ " for idx in comb:\n",
+ " res.append(read_and_parse_data(basedir,dataset,type_list[idx]))\n",
+ " new_df = pd.concat(res)\n",
+ " new_df[new_df.split == \"train\"].to_csv(\"{0}_{1}_train.csv\".format(dataset,\"_\".join([type_list[idx]for idx in comb])),index=None,sep=\"\\t\")\n",
+ " new_df[new_df.split == \"test\"].to_csv(\"{0}_{1}_test.csv\".format(dataset,\"_\".join([type_list[idx]for idx in comb])),index=None,sep=\"\\t\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 22,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>ID</th>\n",
+ " <th>toponym</th>\n",
+ " <th>toponym_context</th>\n",
+ " <th>latitude</th>\n",
+ " <th>longitude</th>\n",
+ " <th>hp_split</th>\n",
+ " <th>split</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>2967171</td>\n",
+ " <td>Ruisseau l'Yvron</td>\n",
+ " <td>Verneuil-l'Étang</td>\n",
+ " <td>48.65648</td>\n",
+ " <td>2.92448</td>\n",
+ " <td>24423</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>2967171</td>\n",
+ " <td>Ruisseau l'Yvron</td>\n",
+ " <td>Bureau de Poste de Noisy Le Sec Principal</td>\n",
+ " <td>48.65648</td>\n",
+ " <td>2.92448</td>\n",
+ " <td>24423</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>2967171</td>\n",
+ " <td>Ruisseau l'Yvron</td>\n",
+ " <td>Maison-Rouge</td>\n",
+ " <td>48.65648</td>\n",
+ " <td>2.92448</td>\n",
+ " <td>24423</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>2967171</td>\n",
+ " <td>Ruisseau l'Yvron</td>\n",
+ " <td>Réau</td>\n",
+ " <td>48.65648</td>\n",
+ " <td>2.92448</td>\n",
+ " <td>24423</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>2967186</td>\n",
+ " <td>Yvette</td>\n",
+ " <td>Maison de la Chimie</td>\n",
+ " <td>48.67032</td>\n",
+ " <td>2.33740</td>\n",
+ " <td>23982</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>...</th>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " <td>...</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>496151</th>\n",
+ " <td>12036176</td>\n",
+ " <td>Hawks Creek Golf Club</td>\n",
+ " <td>Castleberry Elementary School</td>\n",
+ " <td>32.76117</td>\n",
+ " <td>-97.42910</td>\n",
+ " <td>45173</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>496152</th>\n",
+ " <td>12070603</td>\n",
+ " <td>Sugarland</td>\n",
+ " <td>Elsik High School</td>\n",
+ " <td>29.61678</td>\n",
+ " <td>-95.64283</td>\n",
+ " <td>49784</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>496153</th>\n",
+ " <td>12070603</td>\n",
+ " <td>Sugarland</td>\n",
+ " <td>Hedwig Village City Hall</td>\n",
+ " <td>29.61678</td>\n",
+ " <td>-95.64283</td>\n",
+ " <td>49784</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>496154</th>\n",
+ " <td>12070603</td>\n",
+ " <td>Sugarland</td>\n",
+ " <td>Attucks Middle School</td>\n",
+ " <td>29.61678</td>\n",
+ " <td>-95.64283</td>\n",
+ " <td>49784</td>\n",
+ " <td>test</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>496155</th>\n",
+ " <td>12070603</td>\n",
+ " <td>Sugarland</td>\n",
+ " <td>Southside Place</td>\n",
+ " <td>29.61678</td>\n",
+ " <td>-95.64283</td>\n",
+ " <td>49784</td>\n",
+ " <td>train</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "<p>496156 rows × 7 columns</p>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " ID toponym \\\n",
+ "0 2967171 Ruisseau l'Yvron \n",
+ "1 2967171 Ruisseau l'Yvron \n",
+ "2 2967171 Ruisseau l'Yvron \n",
+ "3 2967171 Ruisseau l'Yvron \n",
+ "4 2967186 Yvette \n",
+ "... ... ... \n",
+ "496151 12036176 Hawks Creek Golf Club \n",
+ "496152 12070603 Sugarland \n",
+ "496153 12070603 Sugarland \n",
+ "496154 12070603 Sugarland \n",
+ "496155 12070603 Sugarland \n",
+ "\n",
+ " toponym_context latitude longitude \\\n",
+ "0 Verneuil-l'Étang 48.65648 2.92448 \n",
+ "1 Bureau de Poste de Noisy Le Sec Principal 48.65648 2.92448 \n",
+ "2 Maison-Rouge 48.65648 2.92448 \n",
+ "3 Réau 48.65648 2.92448 \n",
+ "4 Maison de la Chimie 48.67032 2.33740 \n",
+ "... ... ... ... \n",
+ "496151 Castleberry Elementary School 32.76117 -97.42910 \n",
+ "496152 Elsik High School 29.61678 -95.64283 \n",
+ "496153 Hedwig Village City Hall 29.61678 -95.64283 \n",
+ "496154 Attucks Middle School 29.61678 -95.64283 \n",
+ "496155 Southside Place 29.61678 -95.64283 \n",
+ "\n",
+ " hp_split split \n",
+ "0 24423 test \n",
+ "1 24423 train \n",
+ "2 24423 train \n",
+ "3 24423 train \n",
+ "4 23982 train \n",
+ "... ... ... \n",
+ "496151 45173 train \n",
+ "496152 49784 train \n",
+ "496153 49784 test \n",
+ "496154 49784 test \n",
+ "496155 49784 train \n",
+ "\n",
+ "[496156 rows x 7 columns]"
+ ]
+ },
+ "execution_count": 22,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "df_adj"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3.7.5 64-bit",
+ "language": "python",
+ "name": "python37564bitdc8b0e1290e74b85b0e630c435ea2fe8"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.5"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/scripts/embeddingngram.py b/scripts/embeddingngram.py
deleted file mode 100644
index a9773eb..0000000
--- a/scripts/embeddingngram.py
+++ /dev/null
@@ -1,59 +0,0 @@
-#!/usr/bin/env python
-# coding: utf-8
-
-
-from lib.ngram_index import NgramIndex
-from lib.utils_geo import read_geonames
-
-
-
-import pandas as pd
-import numpy as np
-from tqdm import tqdm
-
-
-from tqdm import tqdm
-
-
-
-from gensim.models import Word2Vec
-import logging
-logging.basicConfig(level="INFO")
-
-
-
-df_cooc = pd.read_csv("../data/wikipedia/cooccurrence_ALL.txt",sep="\t")
-df_geo = read_geonames("../data/geonamesData/allCountries.txt")
-
-
-geonames_label = df_geo.name.values.tolist()
-wiki_labels = df_cooc.title.values.tolist()
-p= [wiki_labels.extend(x.split("|")) for x in df_cooc["interlinks"].values]
-
-
-del df_geo
-del df_cooc
-
-N = 5
-
-
-ng = NgramIndex(N)
-p = [ng.split_and_add(x) for x in tqdm(geonames_label)]
-p = [ng.split_and_add(x) for x in tqdm(wiki_labels)]
-ng.save("{0}gramWiki+Geonames_index.json".format(N))
-
-geonames_label.extend(wiki_labels)
-
-class MySentences(object):
- def __init__(self, texts):
- self.texts = texts
-
- def __iter__(self):
- for w in self.texts:
- yield [str(x)for x in ng.encode(w)]
-
-model = Word2Vec(MySentences(geonames_label), size=100, window=5, min_count=1, workers=4,sg=1)
-model.save("embedding{0}gramWiki+Geonames.bin".format(5))
-
-
-
diff --git a/scripts/extract_pages_of_interest.py b/scripts/extract_pages_of_interest.py
new file mode 100644
index 0000000..ef558b9
--- /dev/null
+++ b/scripts/extract_pages_of_interest.py
@@ -0,0 +1,48 @@
+import json
+import gzip
+import argparse
+
+from joblib import Parallel, delayed
+
+# To avoid progressbar issue
+from tqdm import tqdm
+
+
+
+parser = argparse.ArgumentParser()
+parser.add_argument("wikidata_json_dump_filename",help="Wikipedia JSON dump compressed with gzip (*.gz)")
+parser.add_argument("output_filename")
+
+args = parser.parse_args()
+
+# Prepare Output File
+output = open(args.output_filename,'w')
+output.write("{0}\t{1}\t{2}\t{3}\t{4}\t{5}\n".format("ID_WIKIDATA","title","url","latitude","longitude","classes"))
+
+def job(line):
+ line = line.decode("utf-8")
+
+ if not "\"P625\"" in line or not "\"P31\"" in line:
+ return
+ try:
+ data = json.loads(line.strip(",\n"))
+ if "sitelinks" in data and "claims" in data:
+ if "enwiki" in data["sitelinks"] or "frwiki" in data["sitelinks"]:
+ page_available = [i for i in ["en","fr"] if i+"wiki" in data["sitelinks"]]
+ for site in page_available:
+ site = 'en' if 'enwiki' in data["sitelinks"] else 'fr'
+ id_ = data["id"]
+ coords_data = data["claims"]["P625"][0]["mainsnak"]["datavalue"]["value"]
+ title = data["sitelinks"]["{0}wiki".format(site)]["title"]
+ url = "https://{1}.wikipedia.org/wiki/{0}".format(title.replace(" ","_"),site)
+ lat = coords_data["latitude"]
+ lon = coords_data["longitude"]
+ classes_ = ""
+ for claimP31 in data["claims"]["P31"]:
+ classes_ = classes_ + "_"+ str(claimP31["mainsnak"]["datavalue"]["value"]["id"])
+ output.write("{0}\t{1}\t{2}\t{3}\t{4}\t{5}\n".format(id_,title,url,lat,lon,classes_.strip("_")))
+ except Exception: # First Line is "['" and last line is "]'"
+ pass
+
+
+Parallel(n_jobs=8,backend="multiprocessing")(delayed(job)(line)for line in tqdm(gzip.GzipFile(args.wikidata_json_dump_filename),unit_scale=True,unit_divisor=1000))
diff --git a/scripts/generate_cooc_geocoding_dataset.py b/scripts/generate_cooc_geocoding_dataset.py
deleted file mode 100644
index 232f40b..0000000
--- a/scripts/generate_cooc_geocoding_dataset.py
+++ /dev/null
@@ -1,41 +0,0 @@
-import pandas as pd
-import re
-
-####Â TODO NEED TO add ARGPARSE !!!
-def parse_title_wiki(title_wiki):
- """
- Parse Wikipedia title
-
- Parameters
- ----------
- title_wiki : str
- wikipedia title
-
- Returns
- -------
- str
- parsed wikipedia title
- """
- return re.sub("\(.*\)", "", title_wiki).strip().lower()
-
-
-df = pd.read_csv("./cooccurrence_US_FR.txt",sep="\t")
-
-df["interlinks"] = df.interlinks.apply(lambda x : x.split("|"))
-df["interlinks"] = df.interlinks.apply(lambda x : [parse_title_wiki(i) for i in x])
-
-df["title"] = df.title.apply(parse_title_wiki)
-
-def generated_inputs(x):
- output = []
- for interlink in x.interlinks:
- output.append([x.title,interlink,x.longitude,x.latitude])
- return output
-
-output_ = []
-for ix,row in df.iterrows():
- output_.extend(generated_inputs(row))
-
-new_df = pd.DataFrame(output_,columns="name1 name2 longitude latitude".split())
-new_df = new_df.sample(frac=1)
-new_df.to_csv("us_fr_cooc_test.csv",index=False)
\ No newline at end of file
diff --git a/scripts/get_all_adjacency_rel.py b/scripts/get_all_adjacency_rel.py
deleted file mode 100644
index 23382a6..0000000
--- a/scripts/get_all_adjacency_rel.py
+++ /dev/null
@@ -1,88 +0,0 @@
-import pandas as pd, numpy as np
-from numba import njit
-from helpers import read_geonames
-from tqdm import tqdm
-from joblib import Parallel,delayed
-import geopandas as gpd
-from lib.utils_geo import Grid,haversine_pd
-import matplotlib.pyplot as plt
-
-import argparse
-
-parser = argparse.ArgumentParser()
-
-parser.add_argument("geoname_fn")
-parser.add_argument("kilometer_threshold",type=int,default=20)
-parser.add_argument("output_fn_prefix")
-
-args = parser.parse_args("../data/geonamesData/allCountries.txt 20 /home/jacques/ALL_ADJ_224+_".split())
-
-GEONAME_FN = args.geoname_fn
-PREFIX_OUTPUT_FN = args.output_fn_prefix
-KM_THRESHOLD = args.kilometer_threshold
-
-df = read_geonames(GEONAME_FN)
-
-def to_str(list_):
- """
- Return str representation for each value in list_
-
- Parameters
- ----------
- list_ : array
- array
-
- Returns
- -------
- array
- str list
- """
- return list(map(str,list_))
-
-def get_adjacent(geonameid,ids,lon1, lat1, lon2, lat2,threshold):
- """
- Write adjacent entry in geonames for a selected entry
- """
- dist_ = haversine_pd(lon1, lat1, lon2, lat2)
- adj_ids = ids[dist_<threshold]
- out_.write("\n{0},{1},{2},{3}".format(geonameid,"|".join(to_str(adj_ids)),lat2,lon2))
- out_.flush()
-
-
-# WE BUILD a grid over the world map
-# It allows to limit unnecessary calculus thus accelerate the whole process
-world = gpd.read_file("/media/jacques/DATA/GEODATA/WORLD/world.geo.50m.dissolved")
-g = Grid(*world.bounds.values[0],[40,20]) #We build a grid of cell of 40° by 20°
-g.fit_data(world)
-
-# Prepare first output
-first_output_fn = "{1}{0}_cells.csv".format(KM_THRESHOLD,PREFIX_OUTPUT_FN)
-out_ = open(first_output_fn,'w')
-out_.write("geonameid,adjacent_geonameid,latitude,longitude") # HEADER
-out_.flush() # Avoid writing bugs
-
-def get_rels(cells_list):
- for c in tqdm(cells_list):
-
- mask1 = (df.latitude <= c.bottomright_y) & (df.latitude >= c.upperleft_y)
- new_df = df[mask1].copy()
- mask2 = (new_df.longitude >= c.upperleft_x) & (new_df.longitude <= c.bottomright_x)
- new_df = new_df[mask2]
- for ix,row in new_df.iterrows():
- get_adjacent(row.geonameid,new_df.geonameid.values,new_df.longitude,new_df.latitude,row.longitude,row.latitude,KM_THRESHOLD)
- #Parallel(n_jobs=-1,backend="multiprocessing",temp_folder="/home/jacques/temp/")(delayed(get_adjacent)(row.geonameid,new_df.geonameid.values,new_df.longitude,new_df.latitude,row.longitude,row.latitude,KM_THRESHOLD) for ix,row in new_df.iterrows())
-
-world = gpd.read_file(gpd.datasets.get_path('naturalearth_lowres'))
-ax = world.plot(color="white",edgecolor="black")
-for c in g.cells[224:]:
- ax.plot(*c.box_.exterior.xy)
-plt.show()
-get_rels(g.cells[224:]) #~3h
-
-# Prepare second output
-# second_output_fn = "{1}{0}_inter_cells.csv".format(KM_THRESHOLD,PREFIX_OUTPUT_FN)
-# out_ = open(second_output_fn,'w')
-# out_.write("geonameid,adjacent_geonameid,latitude,longitude") # HEADER
-# out_.flush()# Avoid writing bugs
-
-# get_rels(g.inter_cells) 594
diff --git a/scripts/gethealpix.py b/scripts/gethealpix.py
deleted file mode 100644
index 6e572fd..0000000
--- a/scripts/gethealpix.py
+++ /dev/null
@@ -1,32 +0,0 @@
-
-
-import pandas as pd
-
-from tqdm import tqdm
-tqdm.pandas()
-import argparse
-
-import numpy as np
-import healpy
-# convert lat and lon to a healpix code encoding a region, with a given resolution
-def latlon2healpix( lat , lon , res ):
- lat = np.radians(lat)
- lon = np.radians(lon)
- xs = ( np.cos(lat) * np.cos(lon) )#
- ys = ( np.cos(lat) * np.sin(lon) )# -> Sphere coordinates: https://vvvv.org/blog/polar-spherical-and-geographic-coordinates
- zs = ( np.sin(lat) )#
- return healpy.vec2pix( int(res) , xs , ys , zs )
-
-parser = argparse.ArgumentParser()
-parser.add_argument("input_file")
-parser.add_argument("output_file")
-
-args = parser.parse_args()
-
-df = pd.read_csv(args.input_file,sep="\t")
-df["healpix_256"] = df.progress_apply(lambda row:latlon2healpix(lat=row.latitude,lon=row.longitude,res=256),axis=1)
-df["healpix_64"] = df.progress_apply(lambda row:latlon2healpix(lat=row.latitude,lon=row.longitude,res=64),axis=1)
-df["healpix_32"] = df.progress_apply(lambda row:latlon2healpix(lat=row.latitude,lon=row.longitude,res=32),axis=1)
-df["healpix_1"] = df.progress_apply(lambda row:latlon2healpix(lat=row.latitude,lon=row.longitude,res=1),axis=1)
-
-df.to_csv(args.output_file,sep="\t",index=False)
\ No newline at end of file
diff --git a/scripts/randoludo.py b/scripts/rando_ludo_geocoding.py
similarity index 100%
rename from scripts/randoludo.py
rename to scripts/rando_ludo_geocoding.py
diff --git a/templates/pair_topo.html b/templates/pair_topo.html
index 3366643..03094ab 100644
--- a/templates/pair_topo.html
+++ b/templates/pair_topo.html
@@ -47,7 +47,7 @@
color: "red",
fillColor: "#f03",
fillOpacity: 0.5,
- radius: 100000.0
+ radius: 50000.0
}).addTo(mymap);
{% endif %}
--
GitLab