Hi everyone,
Please help us test the first release candidate for scikit-learn 1.9:
pip install scikit-learn==1.9.0rc1
Changelog: https://scikit-learn.org/dev/whats_new/v1.9.html
This release features a few significant changes that requires special
attention:
- narwhals <https://narwhals-dev.github.io/narwhals/> was adopted as a
new dependency to better handle dataframe interoperability. The minimum
supported version is 2.0.1.
- A new config
<https://scikit-learn.org/stable/modules/generated/sklearn.set_config.html>
option was added to control the sparse interface of all sparse objects
that scikit-learn produces. Set
`sklearn.set_config(sparse_interface="sparray")` to output sparse arrays
instead of sparse matrices (current default). The default will change in
a future release.
- SLEP023: Callback API
<https://scikit-learn-enhancement-proposals.readthedocs.io/en/latest/slep023/proposal.html>
has been accepted and this release ships a minimal implementation, with
callback support in LogisticRegression(solver="lbfgs"), Pipeline,
StandardScaler and all *SearchCV classes. The callback API
<https://scikit-learn.org/1.9/callbacks.html> is experimental for now
and might change without deprecation.
Please let us know about any regression, in particular if you maintain a
project with a dependency on
scikit-learn.
Thanks to everyone who contributed to this release!
Jérémie,
on behalf of the scikit-learn maintainer team.
_______________________________________________
scikit-learn mailing list -- [email protected]
To unsubscribe send an email to [email protected]
https://mail.python.org/mailman3//lists/scikit-learn.python.org
Member address: [email protected]
