refactor code according to pandas 1.5.0 and sklearn 1.2.0 releases

feature_engine/imputation/base_imputer.py

@@ -60,9 +60,7 @@ def transform(self, X: pd.DataFrame) -> pd.DataFrame:
         X = self._transform(X)
 
         # Replace missing data with learned parameters
-        X.fillna(value=self.imputer_dict_, inplace=True)
-
-        return X
+        return X.fillna(value=self.imputer_dict_)
 
     def _get_feature_names_in(self, X):
         """Get the names and number of features in the train set (the dataframe

feature_engine/imputation/categorical.py

@@ -134,7 +134,6 @@ def __init__(
         return_object: bool = False,
         ignore_format: bool = False,
     ) -> None:
-
         if imputation_method not in ["missing", "frequent"]:
             raise ValueError(
                 "imputation_method takes only values 'missing' or 'frequent'"
@@ -180,7 +179,6 @@ def fit(self, X: pd.DataFrame, y: Optional[pd.Series] = None):
             self.imputer_dict_ = {var: self.fill_value for var in self.variables_}
 
         elif self.imputation_method == "frequent":
-
             # if imputing only 1 variable:
             if len(self.variables_) == 1:
                 var = self.variables_[0]
@@ -219,7 +217,6 @@ def fit(self, X: pd.DataFrame, y: Optional[pd.Series] = None):
         return self
 
     def transform(self, X: pd.DataFrame) -> pd.DataFrame:
-
         # Frequent category imputation
         if self.imputation_method == "frequent":
             X = super().transform(X)
@@ -230,13 +227,18 @@ def transform(self, X: pd.DataFrame) -> pd.DataFrame:
 
             # if variable is of type category, we need to add the new
             # category, before filling in the nan
+            add_cats = {}
             for variable in self.variables_:
                 if pd.api.types.is_categorical_dtype(X[variable]):
-                    X[variable].cat.add_categories(
-                        self.imputer_dict_[variable], inplace=True
+                    add_cats.update(
+                        {
+                            variable: X[variable].cat.add_categories(
+                                self.imputer_dict_[variable]
+                            )
+                        }
                     )
 
-            X.fillna(self.imputer_dict_, inplace=True)
+            X = X.assign(**add_cats).fillna(self.imputer_dict_)
 
         # add additional step to return variables cast as object
         if self.return_object:

tests/test_encoding/test_encoders/test_similarity_encoder.py

@@ -142,7 +142,7 @@ def test_nan_behaviour_impute(df_enc_big_na):
 def test_nan_behaviour_ignore(df_enc_big_na):
     encoder = StringSimilarityEncoder(missing_values="ignore")
     X = encoder.fit_transform(df_enc_big_na)
-    assert (X.isna().any(1) == df_enc_big_na.isna().any(1)).all()
+    assert (X.isna().any(axis=1) == df_enc_big_na.isna().any(axis=1)).all()
     assert encoder.encoder_dict_ == {
         "var_A": ["B", "D", "G", "A", "C", "E", "F"],
         "var_B": ["A", "D", "B", "G", "C", "E", "F"],

tests/test_imputation/test_categorical_imputer.py

@@ -150,7 +150,6 @@ def test_error_when_imputation_method_not_frequent_or_missing():
 
 
 def test_error_when_variable_contains_multiple_modes(df_na):
-
     msg = "The variable Name contains multiple frequent categories."
     imputer = CategoricalImputer(imputation_method="frequent", variables="Name")
     with pytest.raises(ValueError) as record:
@@ -245,8 +244,9 @@ def test_variables_cast_as_category_missing(df_na):
     X_reference["Name"] = X_reference["Name"].fillna("Missing")
     X_reference["Studies"] = X_reference["Studies"].fillna("Missing")
 
-    X_reference["City"].cat.add_categories("Missing", inplace=True)
-    X_reference["City"] = X_reference["City"].fillna("Missing")
+    X_reference["City"] = (
+        X_reference["City"].cat.add_categories("Missing").fillna("Missing")
+    )
 
     # test fit attributes
     assert imputer.variables_ == ["Name", "City", "Studies"]
@@ -265,7 +265,6 @@ def test_variables_cast_as_category_missing(df_na):
 
 
 def test_variables_cast_as_category_frequent(df_na):
-
     df_na = df_na.copy()
     df_na["City"] = df_na["City"].astype("category")
 

tests/test_selection/test_drop_high_psi_features.py

@@ -266,7 +266,7 @@ def test_split_col_not_included_in_variables(df):
 def test_error_if_na_in_split_col(df):
     """Test an error is raised if the split column contains missing values."""
     data = df.copy()
-    data["var_3"].iloc[15] = np.nan
+    data.iloc[15, data.columns.get_loc("var_3")] = np.nan
 
     transformer = DropHighPSIFeatures(split_col="var_3")
 
@@ -277,7 +277,7 @@ def test_error_if_na_in_split_col(df):
 def test_raise_error_if_na_in_df(df):
     """Test an error is raised when missing values is set to raise."""
     data = df.copy()
-    data["var_3"].iloc[15] = np.nan
+    data.iloc[15, data.columns.get_loc("var_3")] = np.nan
 
     transformer = DropHighPSIFeatures(missing_values="raise")
 
@@ -288,7 +288,7 @@ def test_raise_error_if_na_in_df(df):
 def test_missing_value_ignored(df):
     """Test if PSI are computed when missing values are present in the dataframe."""
     data = df.copy()
-    data["var_3"].iloc[15] = np.nan
+    data.iloc[15, data.columns.get_loc("var_3")] = np.nan
 
     var_col = [col for col in data if "var" in col]
 
@@ -301,7 +301,7 @@ def test_missing_value_ignored(df):
 def test_raise_error_if_inf_in_df(df):
     """Test an error is raised for inf when missing values is set to raise."""
     data = df.copy()
-    data["var_3"].iloc[15] = np.inf
+    data.iloc[15, data.columns.get_loc("var_3")] = np.nan
 
     transformer = DropHighPSIFeatures(missing_values="raise")
 

tests/test_wrappers/test_sklearn_wrapper.py

@@ -1,6 +1,7 @@
 import numpy as np
 import pandas as pd
 import pytest
+from sklearn import __version__ as skl_version
 from sklearn.base import clone
 from sklearn.datasets import fetch_california_housing
 from sklearn.decomposition import PCA
@@ -50,11 +51,21 @@
 ]
 
 
+def _OneHotEncoder(sparse, drop=None, dtype=np.float64) -> OneHotEncoder:
+    """OneHotEncoder sparse argument has been renamed as sparse_output
+    in scikitlearn >=1.2"""
+
+    if skl_version.split(".")[0] == "1" and int(skl_version.split(".")[1]) >= 2:
+        return OneHotEncoder(sparse_output=sparse, drop=drop, dtype=dtype)
+    else:
+        return OneHotEncoder(sparse=sparse, drop=drop, dtype=dtype)
+
+
 @pytest.mark.parametrize(
     "transformer",
     [
         SimpleImputer(),
-        OneHotEncoder(sparse=False),
+        _OneHotEncoder(sparse=False),
         StandardScaler(),
         SelectKBest(),
     ],
@@ -319,7 +330,7 @@ def test_sklearn_ohe_object_one_feature(df_vartypes):
     variables_to_encode = ["Name"]
 
     transformer = SklearnTransformerWrapper(
-        transformer=OneHotEncoder(sparse=False, dtype=np.int64),
+        transformer=_OneHotEncoder(sparse=False, dtype=np.int64),
         variables=variables_to_encode,
     )
 
@@ -341,7 +352,7 @@ def test_sklearn_ohe_object_many_features(df_vartypes):
     variables_to_encode = ["Name", "City"]
 
     transformer = SklearnTransformerWrapper(
-        transformer=OneHotEncoder(sparse=False, dtype=np.int64),
+        transformer=_OneHotEncoder(sparse=False, dtype=np.int64),
         variables=variables_to_encode,
     )
 
@@ -367,7 +378,7 @@ def test_sklearn_ohe_numeric(df_vartypes):
     variables_to_encode = ["Age"]
 
     transformer = SklearnTransformerWrapper(
-        transformer=OneHotEncoder(sparse=False, dtype=np.int64),
+        transformer=_OneHotEncoder(sparse=False, dtype=np.int64),
         variables=variables_to_encode,
     )
 
@@ -387,7 +398,7 @@ def test_sklearn_ohe_numeric(df_vartypes):
 
 def test_sklearn_ohe_all_features(df_vartypes):
     transformer = SklearnTransformerWrapper(
-        transformer=OneHotEncoder(sparse=False, dtype=np.int64)
+        transformer=_OneHotEncoder(sparse=False, dtype=np.int64)
     )
 
     ref = pd.DataFrame(
@@ -443,7 +454,7 @@ def test_sklearn_ohe_with_crossvalidation():
             (
                 "encode_cat",
                 SklearnTransformerWrapper(
-                    transformer=OneHotEncoder(drop="first", sparse=False),
+                    transformer=_OneHotEncoder(drop="first", sparse=False),
                     variables=["AveBedrms_cat"],
                 ),
             ),
@@ -459,7 +470,7 @@ def test_sklearn_ohe_with_crossvalidation():
 
 
 def test_wrap_one_hot_encoder_get_features_name_out(df_vartypes):
-    ohe_wrap = SklearnTransformerWrapper(transformer=OneHotEncoder(sparse=False))
+    ohe_wrap = SklearnTransformerWrapper(transformer=_OneHotEncoder(sparse=False))
     ohe_wrap.fit(df_vartypes)
 
     expected_features_all = [
@@ -541,7 +552,6 @@ def test_error_when_inverse_transform_not_implemented(transformer):
 )
 @pytest.mark.parametrize("transformer", _transformers)
 def test_get_feature_names_out_transformers(varlist, transformer):
-
     X = fetch_california_housing(as_frame=True).frame
     tr_wrap = SklearnTransformerWrapper(transformer=transformer, variables=varlist)
     Xw = tr_wrap.fit_transform(X)
@@ -600,9 +610,8 @@ def test_get_feature_names_out_polynomialfeatures(varlist):
 
 @pytest.mark.parametrize("varlist", [["Name", "City"], None])
 def test_get_feature_names_out_ohe(varlist, df_vartypes):
-
     transformer = SklearnTransformerWrapper(
-        transformer=OneHotEncoder(sparse=False, dtype=np.int64),
+        transformer=_OneHotEncoder(sparse=False, dtype=np.int64),
         variables=varlist,
     )