Skip to content

Commit 4370736

Browse files
committed
DOC Remove future extensions
1 parent 0580391 commit 4370736

File tree

1 file changed

+1
-25
lines changed

1 file changed

+1
-25
lines changed

slep018/proposal.rst

Lines changed: 1 addition & 25 deletions
Original file line numberDiff line numberDiff line change
@@ -45,7 +45,7 @@ in the pipeline::
4545

4646
# X_trans_df is a pandas DataFrame
4747
X_trans_df = num_preprocessor.fit_transform(X_df)
48-
48+
4949
# X_trans_df is again a pandas DataFrame
5050
X_trans_df = num_preprocessor[0].transform(X_df)
5151

@@ -113,30 +113,6 @@ A list of issues discussing Pandas output are: `#14315
113113
<https://github.com/scikit-learn/scikit-learn/pull/20100>`__, and `#23001
114114
<https://github.com/scikit-learn/scikit-learn/issueas/23001>`__.
115115

116-
Future Extensions
117-
-----------------
118-
For information only!
119-
Sparse Data
120-
...........
121-
122-
The Pandas DataFrame is not suitable to provide column names for sparse data
123-
because it has performance issues as shown in `#16772
124-
<https://github.com/scikit-learn/scikit-learn/pull/16772#issuecomment-615423097>`__.
125-
A future extension to this SLEP is to have a ``"pandas_or_namedsparse"`` option.
126-
This option will use a scikit-learn specific sparse container that subclasses
127-
SciPy's sparse matrices. This sparse container includes the sparse data, feature
128-
names and index. This enables pipelines with Vectorizers without performance
129-
issues::
130-
131-
pipe = make_pipeline(
132-
CountVectorizer(),
133-
TfidfTransformer(),
134-
LogisticRegression(solver="liblinear")
135-
)
136-
pipe.set_output(transform="pandas_or_namedsparse")
137-
138-
# feature names for logistic regression
139-
pipe[-1].feature_names_in_
140116

141117
References and Footnotes
142118
------------------------

0 commit comments

Comments
 (0)