[+] Kernel density estimation

src/api.py

@@ -0,0 +1,68 @@
+import json
+
+import matplotlib.pyplot as plt
+import numpy as np
+from parselmouth import Sound
+from scipy.stats import gaussian_kde
+
+from statistics import *
+
+
+Feature = Literal['pitch', 'f1', 'f2', 'f3', 'tilt']
+Gender = Literal['f', 'm']
+
+
+_kde_functions: dict[Feature, dict[Gender, gaussian_kde]] = {}
+
+
+def load_kde():
+    """
+    Load statistical results into kernel density functions
+
+    :return: Kernel density functions for F and M for pitch, f1, f2, f3, tilt
+    """
+    if _kde_functions:
+        return _kde_functions
+
+    data: dict[Feature, dict[Gender, list[float]]] = {**json.loads(Path('results/frequency-data.json').read_text()),
+                                                      **json.loads(Path('results/tilt-data.json').read_text())}
+
+    # Lowercase keys
+    data = {k.lower(): data[k] for k in data}
+
+    # Fit KDE functions
+    for feature in data:
+        _kde_functions[feature] = {}
+        for gender in data[feature]:
+            kde = gaussian_kde(data[feature][gender], 'scott')
+            _kde_functions[feature][gender] = kde
+
+    return _kde_functions
+
+
+def calculate_feature_means(audio: Sound) -> dict[Feature, float]:
+    s = calculate_freq_statistics(calculate_freq_info(audio))
+    return {'pitch': s.pitch.mean, 'f1': s.f1.mean, 'f2': s.f2.mean, 'f3': s.f3.mean, 'tilt': tilt(audio)}
+
+
+def _calculate_fem_prob(feature: Feature, value: float) -> float:
+    """
+    Calculate probability of a feature sounding feminine
+
+    :return: Ratio between 0 and 1
+    """
+    f = load_kde()[feature]['f'].evaluate([value])[0]
+    m = load_kde()[feature]['m'].evaluate([value])[0]
+    return f / (f + m)
+
+
+def calculate_feature_classification(audio: Sound):
+    """
+    Run statistical classification based on kernel density estimation.
+
+    :param audio: Audio
+    :return: Statistical results {'means': {'pitch': ..., 'f1': ...}, 'fem_prob': {'pitch': ..., 'f1': ...}}
+    """
+    means = calculate_feature_means(audio)
+    fem_prob = {feature: _calculate_fem_prob(feature, means[feature]) for feature in means}
+    return {'means': means, 'fem_prob': fem_prob}

src/statistics.py

@@ -280,7 +280,6 @@ def collect_visualize_freq():
     df = pd.DataFrame({headers[i]: f_means[:, i] for i in range(4)})
     dm = pd.DataFrame({headers[i]: m_means[:, i] for i in range(4)})
     args = dict(orient='h', scale='width', inner='quartile', linewidth=0.5)
-    sns.histplot()
     sns.violinplot(data=df, color=COLOR_PINK, **args)
     sns.violinplot(data=dm, color=COLOR_BLUE, **args)
     [c.set_alpha(0.7) for c in ax.collections]
@@ -346,7 +345,7 @@ def collect_visualize_tilt():
     plt.show()
 
     # Write JSON
-    data = {'f': f_means.tolist(), 'm': m_means.tolist()}
+    data = {'tilt': {'f': f_means.tolist(), 'm': m_means.tolist()}}
     Path('results/tilt-data.json').write_text(json.dumps(data), 'utf-8')
 
 
@@ -364,7 +363,7 @@ if __name__ == '__main__':
     # call_id_vox_celeb(combine_id_freq)
 
     # 3. Collect statistics and draw visualizations
-    collect_visualize_freq()
+    # collect_visualize_freq()
 
     ###########
     # 1. Compute and save all the spectral tilt for vox1