Update app.py

app.py

@@ -107,8 +107,13 @@ def preprocess_data(df):
 
     return df
 
-def cluster_data(df):
-    num_clusters = 15  # Set the number of clusters to 15
+def preprocess_csv_data(df):
+    df = df[df['Answer'] == 'Fallback Message shown']
+    df.rename(columns={'Question': 'texts'}, inplace=True)
+    df['texts'] = df['texts'].astype(str)
+    return preprocess_data(df)
+
+def cluster_data(df, num_clusters):
     vectorizer = TfidfVectorizer(stop_words='english')
     X = vectorizer.fit_transform(df['texts'])
 
@@ -134,16 +139,18 @@ def visualize_clusters(df):
 
 def main(file, num_clusters_to_display):
     try:
-        # Check the file extension to determine how to read the file
-        if file.name.endswith('.xlsx'):
+        file_ext = file.name.split('.')[-1].lower()
+        if file_ext == 'xlsx':
             df = pd.read_excel(file)
-        elif file.name.endswith('.csv'):
+            df = preprocess_data(df)
+        elif file_ext == 'csv':
             df = pd.read_csv(file)
+            df = preprocess_csv_data(df)
         else:
-            return "Unsupported file format. Please upload a .xlsx or .csv file."
-        
-        df = preprocess_data(df)
-        df = cluster_data(df)
+            return "Unsupported file format. Please upload an Excel (.xlsx) or CSV (.csv) file."
+
+        num_clusters = 10  # Set the number of clusters
+        df = cluster_data(df, num_clusters)
         visualize_clusters(df)
 
         cluster_sizes = df['Cluster'].value_counts()
@@ -168,7 +175,7 @@ def main(file, num_clusters_to_display):
 interface = gr.Interface(
     fn=main,
     inputs=[
-        gr.File(label="Upload Excel or CSV File (.xlsx, .csv)"),
+        gr.File(label="Upload Excel or CSV File (.xlsx or .csv)"),
         gr.Slider(1, 10, step=1, label="Number of Categories to Display")
     ],
     outputs=gr.File(label="Clustered Data CSV"),