Update clustering.py

clustering.py

@@ -28,6 +28,18 @@ from wordcloud import WordCloud
 
 from concurrent.futures import ThreadPoolExecutor
 
+import logging
+
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
+
+def log_message(message):
+    """"""
+    logging.info(message)
+
 try:
     import torch
     device = 0 if torch.cuda.is_available() else -1
@@ -615,7 +627,7 @@ def perform_clustering(
 
     if not threshold_values:
         threshold_values = np.round(np.linspace(min_eps, max_eps, n), 2)
-        print(threshold_values)
+        log_message(threshold_values)
         # threshold_values = np.linspace(min_eps, max_eps, n)
 
     cluster_assignments = {}
@@ -625,7 +637,7 @@ def perform_clustering(
     most_similar_comments = {}
 
     for distance_threshold in threshold_values:
-        print(distance_threshold)
+        log_message(distance_threshold)
         clustering = AgglomerativeClustering(
             n_clusters=None,
             distance_threshold=distance_threshold,
@@ -695,7 +707,7 @@ def build_sankey_data(
     comments = []
 
     threshold_values = sorted(cluster_assignments.keys())
-    print("build_sankey_data", threshold_values)
+    log_message(f"build_sankey_data {threshold_values}")
     valid_clusters = {}
 
     for threshold in threshold_values: