improved type handling

app.py

@@ -216,11 +216,10 @@ def get_visualizations(beattimes_table: str, cleanedaudio: gr.Audio):
 
     sr, audiodata = cleanedaudio
     
-    segment_metrics = u.compute_segment_metrics(df, sr, audiodata)
 
+    segment_metrics = u.compute_segment_metrics(df, sr, audiodata)
 
 
-    # Normalize audio data from int16 to float in range [-1, 1]
     audiodata = audiodata.astype(np.float32) / 32768.0
 
     # Create figure with secondary y-axes
@@ -438,11 +437,21 @@ def updateBeatsv2(audio:gr.Audio, uploadeddf:gr.File=None)-> go.Figure:
             encoding="utf-8-sig")
         
         # Drop rows where all columns are NaN (empty)
-        beattimes_table = beattimes_table.dropna(how='all')
+        beattimes_table = beattimes_table.dropna()
+        
     
         # Reset the index after dropping rows
         beattimes_table = beattimes_table.reset_index(drop=True)
 
+        # Convert the 'Beattimes' column to float
+        # Handle both string and numeric values in the Beattimes column
+        if beattimes_table['Beattimes'].dtype == 'object':
+            # If strings, replace commas with dots
+            beattimes_table['Beattimes'] = beattimes_table['Beattimes'].str.replace(',', '.').astype(float)
+        else:
+            # If already numeric, just ensure float type
+            beattimes_table['Beattimes'] = beattimes_table['Beattimes'].astype(float)
+
         # Check if the column "Label (S1=0/S2=1)" exists and rename it
         if "Label (S1=0/S2=1)" in beattimes_table.columns:
             beattimes_table = beattimes_table.rename(columns={"Label (S1=0/S2=1)": "Label (S1=1/S2=0)"})

utils.py

@@ -1,13 +1,14 @@
-import librosa
-import numpy as np
-import plotly.graph_objects as go
-from scipy.signal import savgol_filter, find_peaks
 from scipy.signal import butter, filtfilt, find_peaks
-from sklearn.cluster import KMeans
+from scipy.signal import savgol_filter, find_peaks
 from sklearn.preprocessing import StandardScaler
-import pywt
+from sklearn.cluster import KMeans
+import plotly.graph_objects as go
 import pandas as pd
-from scipy.interpolate import interp1d
+import numpy as np
+import librosa
+import pywt
+
+
 
 
 
@@ -183,6 +184,8 @@ def getBeats(audiodata: np.ndarray, sr: int, method='envelope') -> tuple[float,
     """
     # Denoise and normalize
     audiodata, sr = denoise_audio(audiodata, sr)
+    
+    # Normalize to prevent clipping while maintaining relative amplitudes
     cleaned_audio = audiodata / np.max(np.abs(audiodata))
     
     def get_envelope_peaks():
@@ -364,9 +367,11 @@ def plotBeattimes(beattimes: np.ndarray,
     
     # Process and plot primary beat times
     if isinstance(beattimes[0], str):
-        beat_indices = np.round(np.array(beattimes, dtype=float) * sr).astype(int)
+        beat_indices = np.round(np.array([float(bt.replace(',', '.')) for bt in beattimes]) * sr).astype(int)
     else:
         beat_indices = np.round(beattimes * sr).astype(int)
+
+
     beat_indices = beat_indices[beat_indices < len(audiodata)]
     beat_amplitudes = audiodata[beat_indices]
     
@@ -399,7 +404,11 @@ def plotBeattimes(beattimes: np.ndarray,
     
     # Process and plot secondary beat times if provided
     if beattimes2 is not None:
-        beat_indices2 = np.round(beattimes2 * sr).astype(int)
+        if isinstance(beattimes2[0], str):
+            beat_indices2 = np.round(np.array([float(bt.replace(',', '.')) for bt in beattimes2]) * sr).astype(int)
+        else:
+            beat_indices2 = np.round(beattimes2 * sr).astype(int)
+
         beat_indices2 = beat_indices2[beat_indices2 < len(audiodata)]
         beat_amplitudes2 = audiodata[beat_indices2]
         
@@ -482,8 +491,9 @@ def iterate_beat_segments(beat_times, sr, audio):
         # Extract audio segment
         segment = audio[start_sample:end_sample]
         
-        # Analyze segment with beat information
-        segment_metrics.append(segment_analysis(segment, sr, beat_info))
+        # Analyze segment with beat information if not empty
+        if len(segment) > 0:
+            segment_metrics.append(segment_analysis(segment, sr, beat_info))
 
     return segment_metrics
 
@@ -498,6 +508,7 @@ def segment_analysis(segment, sr, s1s2:list):
     Returns:
     - List of computed metrics
     """
+
     # Duration
     duration = len(segment) / sr
     
@@ -576,9 +587,11 @@ def compute_segment_metrics(beattimes: pd.DataFrame, sr: int, audio: np.ndarray)
 
     segment_metrics = iterate_beat_segments(beattimes, sr, audio)
 
+    print("segment_metrics", segment_metrics)
+
     return segment_metrics
 
-def compute_hrv(s1_to_s2, s2_to_s1, sampling_rate=1000):
+def compute_hrv(s1_to_s2, s2_to_s1, sampling_rate):
     """
     Compute Heart Rate Variability with debug statements
     """