From f866e78289247707793710d41a94b26a9eac2ec1 Mon Sep 17 00:00:00 2001
From: Alice BRENON <alice.brenon@ens-lyon.fr>
Date: Wed, 20 Mar 2024 12:11:12 +0100
Subject: [PATCH] Add the function used to reduce the size of ENE events by
 grouping them byArticle (format required by ENEDensity)

---
 GEODE/ENE/Detector.py | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/GEODE/ENE/Detector.py b/GEODE/ENE/Detector.py
index 5e31d67..02ac4df 100644
--- a/GEODE/ENE/Detector.py
+++ b/GEODE/ENE/Detector.py
@@ -23,3 +23,10 @@ def detector(model, corpus):
         for paragraph in paragraphs:
             eventsOfAnnotation(currentPosition, common, model(paragraph))
             currentPosition += len(paragraph)
+
+def byArticle(measures):
+    measures['count'] = 1
+    counts = measures.groupby(articleKey + ['event']).count().reset_index()
+    return counts.pivot(articleKey,
+                        columns='event',
+                        values='count').fillna(0).astype(int).reset_index()
-- 
GitLab