New functions to extract categorical data

utils/statistics.py

+import numpy as np
+
+
+def most_common_value(values):
+    """Get most common value from array"""
+    return np.bincount(values.astype(int)).argmax()
+
+
+def relative_frequency(values, decimals=6, bins=0):
+    """Get the relative frequency of each possible class"""
+    if bins > 0:
+        f = np.round(np.bincount(values.astype(int), minlength=int(bins)) / len(values), decimals)
+    else:
+        f = np.round(np.bincount(values.astype(int)) / len(values), decimals)
+    return f