From 4a16227c4fcac8e478d6a30c1777de20c51f72b2 Mon Sep 17 00:00:00 2001
From: MAFTOUH Mohammed Amine <mohammed-amine.maftouh@imt-atlantique.net>
Date: Sun, 30 Mar 2025 15:22:59 +0000
Subject: [PATCH] Edit mesures_clustering.py
---
clustering/mesures_clustering.py | 33 +++++++++++++++++++++++++++++++-
1 file changed, 32 insertions(+), 1 deletion(-)
diff --git a/clustering/mesures_clustering.py b/clustering/mesures_clustering.py
index 9c6c2e7..4f17e70 100644
--- a/clustering/mesures_clustering.py
+++ b/clustering/mesures_clustering.py
@@ -1,4 +1,5 @@
from sklearn.metrics import silhouette_score
+
def compute_silhouette_scores(reduced_embeddings, labels):
"""
Calcule les scores de silhouette pour différents nombres de clusters.
@@ -8,4 +9,34 @@ def compute_silhouette_scores(reduced_embeddings, labels):
:return: silhouette score
"""
silhouette_avg = silhouette_score(reduced_embeddings, labels)
- return silhouette_avg
\ No newline at end of file
+ return silhouette_avg
+
+
+def compute_dunn_index(reduced_embeddings, labels):
+ """
+ Calcule l'index de Dunn pour évaluer la qualité du clustering.
+
+ :param reduced_embeddings: Matrice des embeddings réduits
+ :param labels: Labels prédits par les algorithmes de clustering
+ :return: Index de Dunn
+ """
+ unique_labels = np.unique(labels)
+ # calcul des distances intra-cluster
+ intra_cluster_distances = []
+ for label in unique_labels:
+ cluster_points = reduced_embeddings[labels == label]
+ if len(cluster_points) > 1:
+ distances = pdist(cluster_points)
+ intra_cluster_distances.append(np.max(distances))
+ #calcul des distances inter-cluster
+ inter_cluster_distances = []
+ for i, label_i in enumerate(unique_labels):
+ for j, label_j in enumerate(unique_labels):
+ if i < j:
+ points_i = reduced_embeddings[labels == label_i]
+ points_j = reduced_embeddings[labels == label_j]
+ distances = pdist(np.vstack([points_i, points_j]))
+ inter_cluster_distances.append(np.min(distances))
+ #calcul de l'index de Dunn
+ dunn_index = np.min(inter_cluster_distances) / np.max(intra_cluster_distances)
+ return dunn_index
\ No newline at end of file
--
GitLab