@@ -106,7 +106,7 @@ original features:
106
106
- Algorithms that create combinations of a fixed number of features, *e.g. *
107
107
``PolynomialFeatures ``, as opposed to all of
108
108
them where there are many. Note that verbosity considerations and
109
- ``verbose_feature_names `` as explained later can apply here.
109
+ ``verbose_feature_names_out `` as explained later can apply here.
110
110
111
111
This proposal talks about how feature names are generated and not how they are
112
112
propagated.
@@ -244,21 +244,22 @@ However, the following examples produce a somewhat redundant feature names::
244
244
Extensions
245
245
##########
246
246
247
- verbose_feature_names
248
- *********************
247
+ verbose_feature_names_out
248
+ *************************
249
+
249
250
To provide more control over feature names, we could add a boolean
250
- ``verbose_feature_names `` constructor argument to certain transformers.
251
+ ``verbose_feature_names_out `` constructor argument to certain transformers.
251
252
The default would reflect the description above, but changes would allow more verbose
252
253
names in some transformers, say having ``StandardScaler `` map ``'age' `` to ``'scale(age)' ``.
253
254
254
- In case of the ``ColumnTransformer `` example above ``verbose_feature_names ``
255
+ In case of the ``ColumnTransformer `` example above ``verbose_feature_names_out ``
255
256
could remove the estimator names, leading to shorter and less redundant names::
256
257
257
258
[model, make, numeric0, ..., numeric100] ->
258
259
make_column_transformer(
259
260
(OneHotEncoder(), ['model', 'make']),
260
261
(PCA(n_components=3), ['numeric0', ..., 'numeric100']),
261
- verbose_feature_names =False
262
+ verbose_feature_names_out =False
262
263
)
263
264
get_feature_names_out(): ['model_100', 'model_200', ...,
264
265
'make_ABC', 'make_XYZ', ...,
0 commit comments