Fix spurious warning from type_of_target when called on estimator.classes_ #31584
+7
−3
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
jeremiedbb
reviewed
Jul 1, 2025
There was a problem hiding this comment.
Thanks for the PR @saskra. Please add a test for _get_response_values in test_response.py to check that no warning is raised now.
Comment on lines
+417
to
+421
| if ( | ||
| not suppress_warning | ||
| and y.shape[0] > 20 | ||
| and classes.shape[0] > round(0.5 * y.shape[0]) | ||
| ): |
There was a problem hiding this comment.
Since the spurious warning is only triggered because we call type_of_target on classes, I'd prefer to add an extra condition to avoid it rather than adding a new parameter to a public function.
Suggested change
| if ( | |
| not suppress_warning | |
| and y.shape[0] > 20 | |
| and classes.shape[0] > round(0.5 * y.shape[0]) | |
| ): | |
| if y.shape[0] > 20 and y.shape[0] > classes.shape[0] > round(0.5 * y.shape[0]): |
With this suggestion, there's an edge case where a y contains only unique values but it's kind of useless for a classification task so a lot of problems would have happen before that, so I wouldn't worry about it.
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Outdated suggestions cannot be applied.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
Fixes #31583
What does this implement/fix? Explain your changes.
This PR suppresses an unintended warning in get_response_values, where type_of_target is called on estimator.classes_. Since classes_ does not represent full sample-level data, this call may spuriously trigger the warning:
"The number of unique classes is greater than 50% of the number of samples."
This is now avoided by passing suppress_warning=True to type_of_target() at this specific location.
This patch is intentionally minimal and does not affect calls to type_of_target that operate on actual sample labels (y, y_true, etc.).
Any other comments?
This was first observed while calibrating classifiers with many classes. Although the dataset was large and well-balanced, the warning appeared due to how classes_ was passed into type_of_target.
Apologies in advance if this is already known or intentional – this is my first contribution here, and I appreciate any feedback or corrections.
Thanks for your time and for maintaining this great library!