Add input validation to _basePCA.inverse_transform #29310
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
13 tasks
jeremiedbb
approved these changes
Jul 25, 2025
There was a problem hiding this comment.
Thanks for the PR @icfaust.
I simplified the validation to only rely on check_array, which deals with empty input.
If the number of components doesn't match, the following operations will fail. With a not as informative error message though, but good enough. I agree that we could have better error messages in that case, but we need to think about it for all estimators, so I leave that for a future issue/PR.
LGTM.
|
@jeremiedbb hey thanks for getting it ready to go! |
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Outdated suggestions cannot be applied.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.
This PR does input checking for PCA's inverse_transform method. If the input array doesn't have the proper number of components, and generic numpy/array_api error is thrown for the dot product for mismatched indices. A check_is_fitted is also added (as is necessary). After this addition, the extracts the necessary checks from self._validate_data, and applies them for the components instead of the features. The check for array namespace is expanded to follow the convention in PCA.transform.
Have added a supporting test.