Fix association model solve fallback#2205
Merged
Merged
Conversation
Contributor
✅MegaLinter analysis: Success
Notices📣 MegaLinter 9.5.0 is out! Discover the new features and security recommendations in the release announcement. (Skip this info by defining See detailed reports in MegaLinter artifacts
|
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.

Summary
LogisticPairwiseAssociationModel's Newton solve fallback backend-neutral by retrying withpinvfor backend runtime/value errors as well as NumPyLinAlgErrorpinvwhen a backend solve returns non-finite steps, and fail explicitly if the fallback is also non-finiteRationale
The association model fit loop uses the active backend's
linalg.solve, but previously only caughtnumpy.linalg.LinAlgError. Non-NumPy backends may signal singular or unsupported linear solves with different exception types, and some backend solve implementations can return non-finite values instead of raising. In both cases, the existing pseudoinverse fallback was bypassed and bad Newton steps could propagate into fitted parameters.Validation
src/pyrecest/utils/association_models.py,tests/test_association_models.py).github.com; PR CI should run the focused and full test matrix.