Scalers: Add test of minmaxscaler when the feature is Null

lucianolorenti · Jun 17, 2024 · 1ad630c · 1ad630c
1 parent 33e8146
commit 1ad630c
Show file tree

Hide file tree

Showing 2 changed files with 52 additions and 3 deletions.
diff --git a/ceruleo/transformation/features/scalers.py b/ceruleo/transformation/features/scalers.py
@@ -187,8 +187,6 @@ def transform(self, X: pd.DataFrame) -> pd.DataFrame:
         try:
             divisor = self.data_max - self.data_min
 
-
-
             mask = np.abs((divisor)) > 1e-25
             X = X.astype(float)
             X.loc[:, mask] = (

diff --git a/tests/test_scalers.py b/tests/test_scalers.py
@@ -3,10 +3,61 @@
 import numpy as np
 import pandas as pd
 
-from ceruleo.transformation.features.scalers import (RobustMinMaxScaler)
+from ceruleo.transformation.features.scalers import (MinMaxScaler, RobustMinMaxScaler)
 from sklearn.preprocessing import RobustScaler
 
 class TestImputers():
+    def test_MinMaxScaler_withNonde(self):
+        scaler = MinMaxScaler(range=(-1, 1), clip=False)
+
+
+        df1 = pd.DataFrame({
+            'a': [None] * 6000,
+            'b': np.random.randn(6000)*5 + 25
+        })
+        scaler.partial_fit(df1)
+
+
+        df2 = pd.DataFrame({
+            'a': [None] * 6000,
+            'b': np.random.randn(6000)*3 + 88
+        })
+        scaler.partial_fit(df2)
+
+        scaled_df1 = scaler.transform(df1)
+        scaled_df2 = scaler.transform(df2)
+        assert np.all(scaled_df1['b'] >= -1) and np.all(scaled_df1['b'] <= 1)
+
+
+    def test_MinMaxScaler(self):
+
+        scaler = MinMaxScaler(range=(-1, 1), clip=False)
+
+
+        df1 = pd.DataFrame({
+            'a': np.random.randn(6000)*5 + 25,
+            'b': np.random.randn(6000)*5 + 25
+        })
+        scaler.partial_fit(df1)
+
+
+        df2 = pd.DataFrame({
+            'a': np.random.randn(6000)*3 + 15,
+            'b': np.random.randn(6000)*3 + 88
+        })
+        scaler.partial_fit(df2)
+
+        scaled_df1 = scaler.transform(df1)
+        scaled_df2 = scaler.transform(df2)
+        assert np.all(scaled_df1['a'] >= -1) and np.all(scaled_df1['a'] <= 1)
+        assert np.all(scaled_df1['b'] >= -1) and np.all(scaled_df1['b'] <= 1)
+        assert np.all(scaled_df2['a'] >= -1) and np.all(scaled_df2['a'] <= 1)
+        assert np.all(scaled_df2['b'] >= -1) and np.all(scaled_df2['b'] <= 1)
+
+
+
+
+
 
     def test_RobustMinMaxScaler(self):