Update visualization.py

visualization.py

@@ -135,6 +135,20 @@ def plot_combined_mse(df, mse_embeddings, mse_posture, mse_voice, title, anomaly
         df['Seconds'] = df['Timecode'].apply(
             lambda x: sum(float(t) * 60 ** i for i, t in enumerate(reversed(x.split(':')))))
 
+    # Ensure the lengths of the DataFrame and MSE values are consistent
+    min_length = min(len(df), len(mse_embeddings), len(mse_posture), len(mse_voice))
+    df = df.iloc[:min_length].copy()
+    mse_embeddings = mse_embeddings[:min_length]
+    mse_posture = mse_posture[:min_length]
+    mse_voice = mse_voice[:min_length]
+
+    # Remove NaN values
+    valid_mask = ~np.isnan(mse_embeddings) & ~np.isnan(mse_posture) & ~np.isnan(mse_voice)
+    df = df[valid_mask]
+    mse_embeddings = mse_embeddings[valid_mask]
+    mse_posture = mse_posture[valid_mask]
+    mse_voice = mse_voice[valid_mask]
+
     # Function to identify continuous segments
     def get_continuous_segments(seconds, values, max_gap=1):
         segments = []
@@ -208,6 +222,7 @@ def plot_combined_mse(df, mse_embeddings, mse_posture, mse_voice, title, anomaly
     plt.close()
     return fig
 
+
 def plot_mse_histogram(mse_values, title, anomaly_threshold, color='blue'):
     plt.figure(figsize=(16, 3), dpi=300)
     fig, ax = plt.subplots(figsize=(16, 3))