DEPR deprecate n_iter in MiniBatchSparsePCA#23726
DEPR deprecate n_iter in MiniBatchSparsePCA#23726glemaitre merged 10 commits intoscikit-learn:mainfrom
Conversation
jeremiedbb
left a comment
There was a problem hiding this comment.
I'm fine with considering it a bug in the deprecation cycle of MiniBatchDictionaryLearning , thus targeting 1.1 rather than 1.2 to keep both in sync.
Co-authored-by: Jérémie du Boisberranger <34657725+jeremiedbb@users.noreply.github.com>
|
I also move the deprecation warning raise in |
test_spca_early_stopping
Co-authored-by: Jérémie du Boisberranger <34657725+jeremiedbb@users.noreply.github.com>
| To disable convergence detection based on cost function, set | ||
| `max_no_improvement` to `None`. | ||
|
|
||
| .. versionadded:: 1.1 |
There was a problem hiding this comment.
In retrospect it's a bit weird to introduce a new parameter in a bugfix release.
Maybe we should instead move this PR to target 1.2 and update the previous deprecation message for the MB sparse PCA to 1.4 instead of 1.3 so that both the MB dict learning and sparse PCA variant are actually removed in 1.4 at the same time to ease maintenance.
Co-authored-by: Jérémie du Boisberranger <34657725+jeremiedbb@users.noreply.github.com> Co-authored-by: jeremie du boisberranger <jeremiedbb@yahoo.fr> Co-authored-by: Olivier Grisel <olivier.grisel@ensta.org>
This PR is deprecating
n_iterand introducingmax_iter,tol, andmax_no_improvementparameters to be consistent withMiniBatchDictionaryLearning.We should do this step in version 1.1 since some code will have to be removed and
n_iterwould not be available inMiniBatchDictionaryLearningrequired by theMiniBatchSparsePCA.We should therefore backport this fix in 1.1.2.