Hi there, We're calling a vote <https://github.com/scikit-learn/enhancement_proposals/pull/65> on SLEP006 - Metadata Routing, which is available here <https://scikit-learn-enhancement-proposals.readthedocs.io/en/latest/slep006/proposal.html>, and detailed past discussions can be found under these PRs <https://github.com/scikit-learn/enhancement_proposals/pulls?q=is%3Apr++slep006+> and these issues <https://github.com/scikit-learn/enhancement_proposals/issues?q=is%3Aissue+slep006> .
The current proposed implementation is drafted under this pull request <https://github.com/scikit-learn/scikit-learn/pull/22083> where you can find a rendered version of the user guide here <https://171332-843222-gh.circle-artifacts.com/0/doc/metadata_routing.html> and a rendered version of the developer API (only applicable to third party developers and people who write custom estimators) here <https://171332-843222-gh.circle-artifacts.com/0/doc/auto_examples/plot_metadata_routing.html>. These are found under `metadata_routing.rst` and `plot_metadata_routing.py` under the aforementioned PR <https://github.com/scikit-learn/scikit-learn/pull/22083> respectively. Note that this PR does NOT touch scorers, splitters, or any of the estimators or meta-estimators in scikit-learn. It implements the machinery in `BaseEstimator` for *consumers* only. The PR is also not targeted to `main`, and instead it's to be merged into the `sample-props` branch on the main repo, with follow-up PRs to complete the implementation before merging into `main`. Please leave your votes here in the PR <https://github.com/scikit-learn/enhancement_proposals/pull/65>, and comment on the PR, here on the mailing list, or open new PRs/issues in the enhancement proposals repo or the main repo for specifics if you think further discussion is required. Note that this vote is not to merge the implementation PR, and is rather to accept the SLEP, and the SLEP does NOT discuss implementation details and the API for us and third party developers; but we're more than happy to discuss the implementation and the API during this voting round. We also plan to send out a survey asking third party developers for their opinion of our proposed developer API, parallel to this call for vote. This vote will close on February 21, 2022, and per our governance model it will pass if 2/3 of the cast vote from core developers are in favor, with no quorum/minimum participation requirements. Regards, Adrin
_______________________________________________ scikit-learn mailing list scikit-learn@python.org https://mail.python.org/mailman/listinfo/scikit-learn
