DOC improve doc for _check_n_features and _check_feature_names (scikit-learn#31585)

Co-authored-by: Jérémie du Boisberranger <jeremie@probabl.ai>

Author: VirenPassi

sklearn/preprocessing/_function_transformer.py

@@ -16,9 +16,7 @@
 from ..utils.metaestimators import available_if
 from ..utils.validation import (
     _allclose_dense_sparse,
-    _check_feature_names,
     _check_feature_names_in,
-    _check_n_features,
     _get_feature_names,
     _is_pandas_df,
     _is_polars_df,
@@ -178,17 +176,6 @@ def __init__(
         self.kw_args = kw_args
         self.inv_kw_args = inv_kw_args
 
-    def _check_input(self, X, *, reset):
-        if self.validate:
-            return validate_data(self, X, accept_sparse=self.accept_sparse, reset=reset)
-        elif reset:
-            # Set feature_names_in_ and n_features_in_ even if validate=False
-            # We run this only when reset==True to store the attributes but not
-            # validate them, because validate=False
-            _check_n_features(self, X, reset=reset)
-            _check_feature_names(self, X, reset=reset)
-        return X
-
     def _check_inverse_transform(self, X):
         """Check that func and inverse_func are the inverse."""
         idx_selected = slice(None, None, max(1, X.shape[0] // 100))
@@ -240,7 +227,13 @@ def fit(self, X, y=None):
         self : object
             FunctionTransformer class instance.
         """
-        X = self._check_input(X, reset=True)
+        X = validate_data(
+            self,
+            X,
+            reset=True,
+            accept_sparse=self.accept_sparse,
+            skip_check_array=not self.validate,
+        )
         if self.check_inverse and not (self.func is None or self.inverse_func is None):
             self._check_inverse_transform(X)
         return self
@@ -259,7 +252,9 @@ def transform(self, X):
         X_out : array-like, shape (n_samples, n_features)
             Transformed input.
         """
-        X = self._check_input(X, reset=False)
+        if self.validate:
+            X = validate_data(self, X, reset=False, accept_sparse=self.accept_sparse)
+
         out = self._transform(X, func=self.func, kw_args=self.kw_args)
         output_config = _get_output_config("transform", self)["dense"]
 

sklearn/utils/validation.py

@@ -2723,6 +2723,10 @@ def _check_feature_names(estimator, X, *, reset):
         Moved from :class:`~sklearn.base.BaseEstimator` to
         :mod:`sklearn.utils.validation`.
 
+    .. note::
+        To only check feature names without conducting a full data validation, prefer
+        using `validate_data(..., skip_check_array=True)` if possible.
+
     Parameters
     ----------
     estimator : estimator instance
@@ -2733,8 +2737,10 @@ def _check_feature_names(estimator, X, *, reset):
 
     reset : bool
         Whether to reset the `feature_names_in_` attribute.
+        If True, resets the `feature_names_in_` attribute as inferred from `X`.
         If False, the input will be checked for consistency with
         feature names of data provided when reset was last True.
+
         .. note::
            It is recommended to call `reset=True` in `fit` and in the first
            call to `partial_fit`. All other methods that validate `X`
@@ -2810,6 +2816,10 @@ def add_names(names):
 def _check_n_features(estimator, X, reset):
     """Set the `n_features_in_` attribute, or check against it on an estimator.
 
+    .. note::
+        To only check n_features without conducting a full data validation, prefer
+        using `validate_data(..., skip_check_array=True)` if possible.
+
     .. versionchanged:: 1.6
         Moved from :class:`~sklearn.base.BaseEstimator` to
         :mod:`~sklearn.utils.validation`.
@@ -2823,12 +2833,14 @@ def _check_n_features(estimator, X, reset):
         The input samples.
 
     reset : bool
+        Whether to reset the `n_features_in_` attribute.
         If True, the `n_features_in_` attribute is set to `X.shape[1]`.
         If False and the attribute exists, then check that it is equal to
         `X.shape[1]`. If False and the attribute does *not* exist, then
         the check is skipped.
+
         .. note::
-           It is recommended to call reset=True in `fit` and in the first
+           It is recommended to call `reset=True` in `fit` and in the first
            call to `partial_fit`. All other methods that validate `X`
            should set `reset=False`.
     """