From 423c26c3393fa8d208edf62db5e8c30c0ccc94d2 Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Fri, 20 May 2022 15:42:33 +0200
Subject: [PATCH] Added script to print most ambiguous vectors from w2v
generated from lexicon
---
findMostAmbiguousFromW2v.py | 27 +++++++++++++++++++++++++++
1 file changed, 27 insertions(+)
create mode 100755 findMostAmbiguousFromW2v.py
diff --git a/findMostAmbiguousFromW2v.py b/findMostAmbiguousFromW2v.py
new file mode 100755
index 0000000..122217f
--- /dev/null
+++ b/findMostAmbiguousFromW2v.py
@@ -0,0 +1,27 @@
+#! /usr/bin/env python3
+
+# Given a .w2v file, that has been produced by macaon_data/script/lefff2W2V.py, list the most ambiguous words.
+
+import sys
+
+def formatProba(f) :
+ if f > 0.0 :
+ return "\033[1m%.2f\033[0m"%f
+ return "%.2f"%f
+
+if __name__ == "__main__" :
+ wordList = []
+
+ for line in open(sys.argv[1], "r") :
+ line = line.strip()
+ splited = line.split()
+ if len(splited) == 2 :
+ continue
+ word = splited[0]
+ vec = [float(c) for c in splited[1:]]
+ maxProba = max(vec)
+ wordList.append((maxProba, " ".join([formatProba(f) for f in vec]+[word])))
+
+wordList.sort()
+print("\n".join([e[1] for e in wordList if e[0] < 1.0]))
+
--
GitLab