lecrapaud 0.8.0__tar.gz → 0.8.1__tar.gz

{lecrapaud-0.8.0 → lecrapaud-0.8.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: lecrapaud
-Version: 0.8.0
+Version: 0.8.1
 Summary: Framework for machine and deep learning, with regression, classification and time series analysis
 License: Apache License
 Author: Pierre H. Gallet

{lecrapaud-0.8.0 → lecrapaud-0.8.1}/lecrapaud/feature_engineering.py

@@ -122,41 +122,8 @@ class FeatureEngineeringEngine:
         # Cyclic encode dates
         self.data = self.cyclic_encode_date()
 
-        # Remove constant columns if training (feature selection methods need independent features)
-        if self.for_training:
-            self.data = self.remove_constant_columns()
-
         return self.data
 
-    def remove_constant_columns(
-        self,
-        threshold: float = 0.99,
-    ) -> pd.DataFrame:
-        """
-            Removes constant or almost constant columns from a DataFrame.
-
-        Parameters:
-            threshold (float): Max proportion for a single value (default 0.99 = 99%).
-
-        Returns:
-            pd.DataFrame: Cleaned DataFrame.
-        """
-
-        to_drop = []
-
-        for col in self.data.columns:
-            value_counts = self.data[col].value_counts(dropna=False, normalize=True)
-            if value_counts.empty:
-                to_drop.append(col)
-            elif value_counts.iloc[0] >= threshold:
-                to_drop.append(col)
-
-        if to_drop:
-            logger.info(f"🔍 Removed {len(to_drop)} constant/almost constant columns:")
-            logger.info(to_drop)
-
-        return self.data.drop(columns=to_drop, errors="ignore")
-
     def cyclic_encode_date(self) -> pd.DataFrame:
         """
         Adds cyclic (sine and cosine) encoding for common date parts: day of week, day of month, and month.

{lecrapaud-0.8.0 → lecrapaud-0.8.1}/lecrapaud/feature_selection.py

@@ -138,6 +138,7 @@ class FeatureSelectionEngine:
         # Let's start by removing extremly correlated features
         # This is needed to reduce nb of feature but also for methods such as anova or chi2 that requires independent features
         # TODO: we could also remove low variance features
+        self.X = self.remove_constant_columns()
         features_uncorrelated, features_correlated = self.remove_correlated_features(
             90, vizualize=False
         )
@@ -352,6 +353,35 @@ class FeatureSelectionEngine:
     # Remove correlation
     # ------------------
 
+    def remove_constant_columns(
+        self,
+        threshold: float = 0.99,
+    ) -> pd.DataFrame:
+        """
+            Removes constant or almost constant columns from a DataFrame.
+
+        Parameters:
+            threshold (float): Max proportion for a single value (default 0.99 = 99%).
+
+        Returns:
+            pd.DataFrame: Cleaned DataFrame.
+        """
+
+        to_drop = []
+
+        for col in self.data.columns:
+            value_counts = self.data[col].value_counts(dropna=False, normalize=True)
+            if value_counts.empty:
+                to_drop.append(col)
+            elif value_counts.iloc[0] >= threshold:
+                to_drop.append(col)
+
+        if to_drop:
+            logger.info(f"🔍 Removed {len(to_drop)} constant/almost constant columns:")
+            logger.info(to_drop)
+
+        return self.data.drop(columns=to_drop, errors="ignore")
+
     def remove_correlated_features(self, corr_threshold: int, vizualize: bool = False):
         X = self.X
         features = X.columns

{lecrapaud-0.8.0 → lecrapaud-0.8.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "lecrapaud"
-version = "0.8.0"
+version = "0.8.1"
 description = "Framework for machine and deep learning, with regression, classification and time series analysis"
 authors = [
     {name = "Pierre H. Gallet"}