FIX - Make SelectCols and DropCols implement sklearn convention (skrub-data#1813)

Author: auguste-probabl

CHANGES.rst

@@ -17,6 +17,10 @@ Changes
 
 Bugfixes
 --------
+- :class:`DropCols` and :class:`SelectCols:` attributes were renamed to end
+  with an underscore, in order to follow a scikit-learn convention which is
+  used to determine if an estimator is fitted. :pr:`1813` by :user:`Auguste
+  Baum <auguste-probabl>`.
 
 Release 0.7.0
 =============

skrub/_select_cols.py

@@ -63,7 +63,7 @@ def fit(self, X, y=None):
         SelectCols
             The transformer itself.
         """
-        self._columns = s.make_selector(self.cols).expand(X)
+        self.columns_ = s.make_selector(self.cols).expand(X)
         return self
 
     def transform(self, X):
@@ -80,7 +80,7 @@ def transform(self, X):
             The input DataFrame ``X`` after selecting only the columns listed
             in ``self.cols`` (in the provided order).
         """
-        return s.select(X, self._columns)
+        return s.select(X, self.columns_)
 
     def get_feature_names_out(self, input_features=None):
         """Get output feature names for transformation.
@@ -95,8 +95,8 @@ def get_feature_names_out(self, input_features=None):
         feature_names_out : ndarray of str objects
             Transformed feature names.
         """
-        check_is_fitted(self, "_columns")
-        return self._columns
+        check_is_fitted(self, "columns_")
+        return self.columns_
 
 
 class DropCols(TransformerMixin, BaseEstimator):
@@ -159,8 +159,8 @@ def fit(self, X, y=None):
             The transformer itself.
         """
         selector = s.make_selector(self.cols)
-        self._kept_cols = (~selector).expand(X)
-        self._dropped_cols = selector.expand(X)
+        self.kept_cols_ = (~selector).expand(X)
+        self.dropped_cols_ = selector.expand(X)
         return self
 
     def transform(self, X):
@@ -177,7 +177,7 @@ def transform(self, X):
             The input DataFrame ``X`` after dropping the columns listed in
             ``self.cols``.
         """
-        return s.select(X, s.make_selector(self._kept_cols))
+        return s.select(X, s.make_selector(self.kept_cols_))
 
     def get_feature_names_out(self, input_features=None):
         """Get output feature names for transformation.
@@ -192,8 +192,8 @@ def get_feature_names_out(self, input_features=None):
         feature_names_out : ndarray of str objects
             Transformed feature names.
         """
-        check_is_fitted(self, "_kept_cols")
-        return self._kept_cols
+        check_is_fitted(self, "kept_cols_")
+        return self.kept_cols_
 
 
 class Drop(SingleColumnTransformer):