From 0ccb0c4979221ebddeb9d0c604eb7561f4cf842f Mon Sep 17 00:00:00 2001
From: Fize Jacques <jacques.fize@cirad.fr>
Date: Tue, 2 Feb 2021 10:44:03 +0100
Subject: [PATCH] add multithread run_eval script + debug split graph error
---
evalNE_script.py | 5 +++-
run_eval_par.py | 59 ++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 63 insertions(+), 1 deletion(-)
create mode 100644 run_eval_par.py
diff --git a/evalNE_script.py b/evalNE_script.py
index 7c1ceef..7f29050 100644
--- a/evalNE_script.py
+++ b/evalNE_script.py
@@ -40,7 +40,10 @@ log("Density "+ str(len(G)/len(list(G.edges()))))
log("Building link prediction dataset...")
# Create an evaluator and generate train/test edge split
traintest_split = LPEvalSplit()
-traintest_split.compute_splits(G,split_alg="fast",train_frac=0.6,fe_ratio=1)
+try:
+ traintest_split.compute_splits(G,split_alg="fast",train_frac=0.6,fe_ratio=1)
+except ValueError:
+ traintest_split.compute_splits(G, split_alg="spanning_tree", train_frac=0.6, fe_ratio=1)
nee = LPEvaluator(traintest_split)
log("Dataset Built !")
diff --git a/run_eval_par.py b/run_eval_par.py
new file mode 100644
index 0000000..16e4a1f
--- /dev/null
+++ b/run_eval_par.py
@@ -0,0 +1,59 @@
+# coding = utf-8
+
+import glob
+import subprocess
+from lib.helpers import parse_evalne_output
+from lib.utils import load_edgelist
+import os
+import pandas as pd
+from tqdm import tqdm
+import networkx as nx
+from joblib import Parallel,delayed
+
+import argparse
+
+parser = argparse.ArgumentParser()
+parser.add_argument("dataset_dir")
+parser.add_argument("output_filename")
+parser.add_argument("-f", "--format", default="gexf", choices=["gexf", "gml", "txt"])
+
+args = parser.parse_args()
+fns = sorted(glob.glob(args.dataset_dir + "/*." + args.format))
+
+def run_eval(fn):
+ command = "python evalNE_script.py {0} -f {1} -n".format(fn, args.format).split()
+ output = subprocess.run(command)
+ if not output.returncode == 0:
+ print("Error! for the command :", " ".join(command))
+
+all_res = []
+
+# Run link prediction
+Parallel(n_jobs=4,backend="multiprocessing")(delayed(run_eval)(fn) for fn in tqdm(fns))
+
+pbar = tqdm(fns)
+for fn in pbar:
+ pbar.set_description("compile eval from "+ fn)
+
+ if os.path.exists(fn + "_results_lp"):
+ df_results = parse_evalne_output(open(fn + "_results_lp").read())
+ name = os.path.basename(fn)
+ G = None
+ if args.format == "edgelist":
+ G = load_edgelist(path=fn)
+ elif args.format == "gml":
+ G = nx.read_gml(fn)
+ else:
+ G = nx.read_gexf(fn)
+
+ top10node = pd.DataFrame(list(G.degree()), columns="node degree".split()).sort_values("degree",
+ ascending=False).head(10).node.values
+ df_results["nb_edge"] = G.size()
+ df_results["transitivity"] = nx.transitivity(G)
+ df_results["density"] = nx.density(G)
+ df_results["top10_node"] = "|".join(top10node)
+ df_results["size"] = len(G)
+ df_results["filename"] = name
+ all_res.append(df_results)
+
+pd.concat(all_res).to_csv(args.output_filename, sep="\t", index=False)
--
GitLab