def test_get_feature_names_out_polynomialfeatures(varlist):
X = fetch_california_housing(as_frame=True).frame
tr_wrap = SklearnTransformerWrapper(transformer=PolynomialFeatures(),
variables=varlist)
Xw = tr_wrap.fit_transform(X)
assert Xw.columns.tolist() == tr_wrap.get_feature_names_out()
if varlist is not None:
output_feat = [
"1",
"MedInc",
"HouseAge",
"AveRooms",
"AveBedrms",
"MedInc^2",
"MedInc HouseAge",
"MedInc AveRooms",
"MedInc AveBedrms",
"HouseAge^2",
"HouseAge AveRooms",
"HouseAge AveBedrms",
"AveRooms^2",
"AveRooms AveBedrms",
"AveBedrms^2",
]
assert output_feat == tr_wrap.get_feature_names_out(varlist)
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)
assert Xw.columns.to_list() == tr_wrap.get_feature_names_out()
assert Xw.columns.to_list() == tr_wrap.get_feature_names_out(
["MedInc", "HouseAge"])
def test_get_feature_names_out_selectors(varlist, transformer):
X = fetch_california_housing(as_frame=True).frame
y = X["MedHouseVal"]
X = X.drop(["MedHouseVal"], axis=1)
tr_wrap = SklearnTransformerWrapper(transformer=transformer,
variables=varlist)
Xw = tr_wrap.fit_transform(X, y)
assert Xw.columns.to_list() == tr_wrap.get_feature_names_out()
assert Xw.columns.to_list() == tr_wrap.get_feature_names_out(
["MedInc", "HouseAge"])
def test_get_feature_names_out_ohe(varlist, df_vartypes):
transformer = SklearnTransformerWrapper(
transformer=OneHotEncoder(sparse=False, dtype=np.int64),
variables=varlist,
)
df_tr = transformer.fit_transform(df_vartypes)
assert df_tr.columns.to_list() == transformer.get_feature_names_out()
if varlist is not None:
output_feat = [
"Name_jack",
"Name_krish",
"Name_nick",
"Name_tom",
"City_Bristol",
"City_Liverpool",
"City_London",
"City_Manchester",
]
assert output_feat == transformer.get_feature_names_out(varlist)