Add the code used to annotate ENE and export labels while we're at it

GEODE/ENE/Detector.py

+from GEODE.Metadata import articleKey, fromKey, toKey
+import pandas
+import sys
+
+columns = articleKey + ['event', 'position', 'size', 'totalSize']
+
+def eventsOfAnnotation(start, common, annotated):
+    rows = pandas.DataFrame([{**common,
+                              'event': span.label_,
+                              'position': start + span.start_char,
+                              'size': span.end_char - span.start_char}
+                             for span in annotated.spans['sc']])
+    if len(rows):
+        rows[columns].to_csv(sys.stdout, sep='\t', index=False, header=False)
+
+def detector(model, corpus):
+    print(*columns, sep='\t')
+    for text in corpus.get_all():
+        paragraphs = text['content'].split('\n\n')
+        common = {**fromKey(toKey(text)),
+                  'totalSize': sum([len(p) for p in paragraphs])}
+        currentPosition = 0
+        for paragraph in paragraphs:
+            eventsOfAnnotation(currentPosition, common, model(paragraph))
+            currentPosition += len(paragraph)

GEODE/ENE/__init__.py

+ene_labels = [
+    "Domain-mark",
+    "Head",
+    "Latlong",
+    "NC-Spatial",
+    "NC-Person",
+    "NP-Misc",
+    "NP-Person",
+    "NP-Spatial",
+    "Relation",
+    "ENE-Spatial",
+    "ENE-Person",
+    "ENE-Misc"
+        ]

GEODE/ENE/__main__.py

+from GEODE.ENE.Detector import detector
+from GEODE.Store import corpus
+import spacy
+import sys
+
+detector(spacy.load(sys.argv[1]), corpus(sys.argv[2]))

guix.scm

-(use-modules ((gnu packages machine-learning) #:select (python-pytorch python-scikit-learn))
+(use-modules ((gnu packages machine-learning) #:select (python-pytorch python-scikit-learn python-spacy))
              ((gnu packages python-science) #:select (python-pandas))
              ((gnu packages python-xyz) #:select (python-matplotlib
                                                   python-nltk
@@ -23,7 +23,8 @@
       (list python-matplotlib
             python-pandas
             python-scikit-learn
-            python-seaborn))
+            python-seaborn
+            python-spacy))
     (arguments
      (list #:tests? #f))
     (home-page "https://gitlab.liris.cnrs.fr/geode/geopyck")

requirements.txt

+matplotlib
 pandas 
 scikit-learn
-matplotlib
 seaborn
+spacy