MNT Add validation for parameter alphas in RidgeCV
#21606
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jjerphan
reviewed
Nov 18, 2021
sklearn/linear_model/_ridge.py
Outdated
Comment on lines
1847
to
1848
| if isinstance(self.alphas, (np.ndarray, list, tuple)): | ||
| n_alphas = 1 if np.ndim(self.alphas) == 0 else len(self.alphas) |
There was a problem hiding this comment.
As discussed with @ArturoAmorQ , this provides correct parameter validation to all sub-classes of _RidgeGCV.
Yet, in order for the validation to be run once, we need, for each subclass:
- to move those validation at the beginning of
fit - to move the core of
_fitin a dedicated private method (_fit) which does all but the validation - to move other validation steps as well (?)
- to check that the validation is only done once.
For us, this small refactoring has be done in another PR.
cc @glemaitre: we need your authority on the best way to proceed.
There was a problem hiding this comment.
Yes, this is indeed what we are leaning towards as a solution. We already applied such a pattern with @jeremiedbb: validating in fit that will call _fit where the processing is done with no validation.
b95f434 to
a781866
Compare
glemaitre
reviewed
Nov 22, 2021
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
glemaitre
reviewed
Dec 1, 2021
glemaitre
reviewed
Dec 1, 2021
Co-authored-by: Julien Jerphanion <git@jjerphan.xyz>
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
added 2 commits
December 1, 2021 14:17
…earn into validations_ridgecv
jjerphan
reviewed
Dec 1, 2021
Co-authored-by: Julien Jerphanion <git@jjerphan.xyz>
jjerphan
approved these changes
Dec 1, 2021
|
@jjerphan on the contrary, thank you! And also thank you @glemaitre for your time and feedback! |
glemaitre
reviewed
Dec 1, 2021
added 4 commits
December 1, 2021 17:21
|
Thanks @ArturoAmorQ |
sheikhmunim
added a commit
to sheikhmunim/scikit-learn
that referenced
this pull request
Dec 2, 2021
MNT Add validation for parameter `alphas` in `RidgeCV` (scikit-learn#21606)
41 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Reference Issues/PRs
Solves a part of the Issue #20724 as mentioned in this comment.
What does this implement/fix? Explain your changes.
Use the helper function
check_scalarfromsklearn.utilsto validate the scalar parameteralphasand make sure to get consistent error types and messages inlinear_model.RidgeCV.In the case that
alphasis a list, anp.ndarrayor a tuple, the check iterates through the elements and gives feedback on the position of the first parameter that fails the check.Any other comments?
This PR is a first approach to implement the validation for the scalar parameter
alphas. Making the process more efficient whenfitis called several times (large number of folds in the cv or large number of points in a grid search) will be addressed in a future PR.