gini can be negative by your code

lifetime_value/metrics.py

@@ -18,7 +18,7 @@
 
 import numpy as np
 import pandas as pd
-from sklearn import metrics
+import sklearn.metrics as sk_metrics
 
 
 def cumulative_true(
@@ -62,11 +62,11 @@ def gini_from_gain(df: pd.DataFrame) -> pd.DataFrame:
 
 
 def _normalized_rmse(y_true, y_pred):
-  return np.sqrt(metrics.mean_squared_error(y_true, y_pred)) / y_true.mean()
+  return np.sqrt(sk_metrics.mean_squared_error(y_true, y_pred)) / y_true.mean()
 
 
 def _normalized_mae(y_true, y_pred):
-  return metrics.mean_absolute_error(y_true, y_pred) / y_true.mean()
+  return sk_metrics.mean_absolute_error(y_true, y_pred) / y_true.mean()
 
 
 def _aggregate_fn(df):

lifetime_value/metrics_test.py

@@ -14,7 +14,13 @@
 # ============================================================================
 # Lint as: python3
 # Dependency imports
-
+import sys
+import os
+import io
+current_path = os.getcwd()
+sys.path.append(os.getcwd())
+sys.path.append(os.path.join(current_path, "lifetime_value"))
+print(sys.path)
 from lifetime_value import metrics
 import numpy as np
 import pandas as pd
@@ -61,6 +67,30 @@ def test_decile_stats(self):
             rtol=1e-2,
             atol=1000))
 
+  def test_gini_negative(self):
+    test_df2 = """a,0.1,0.115
+b,0.1,0.112
+c,0.1,0.1151
+d,0.9,0.01"""
+    df = pd.read_csv(io.StringIO(test_df2), header=None)
+    column_names = ['uid', 'label1', 'pred_scores']
+    df.columns = column_names
+    df = df.sort_values(by='label1', ascending=False)
+    print(f"test_df: {df.head(10)}")
+
+
+    total_value = np.sum(df['label1'])
+    cumulative_true = np.cumsum(df['label1']) / total_value
+    gain_model = metrics.cumulative_true(
+        df['label1'], df['pred_scores'])
+    gain = pd.DataFrame({
+        'ground_truth': cumulative_true,
+        'random_model': gain_model
+    })
+    gini = metrics.gini_from_gain(gain)
+    print(f"test_gini: {gini.head(10)}")
+    assert gini.loc['random_model', 'raw'] < 0
+
 
 if __name__ == '__main__':
   unittest.main()