Repository navigation
Fix numerical baseline checks across numerical environments - #15
Merged
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
The numerical baseline checks fail with current rdrobust calculations because their bandwidth tolerances are tighter than the observed numerical variation. Allow a relative tolerance of 1e-6 for the historical bandwidth baselines in Python and R, and align the affected R p-value comparison with the Python relative tolerance. Retain the saved values and all other checks.
Implementation
Checks
Python: all three numerical contracts pass; source and wheel builds succeed. R: numerical contracts pass, source build succeeds, and R CMD check --no-manual returns Status: OK. Separate comparisons of the cutoff and pooled estimates, bandwidths, p-values, and sample counts agree with direct rdrobust calls in Python and R at a relative tolerance of 1e-12 when options match.
Notes
The change is limited to the two existing numerical test files. Package implementation and statistical formulas are unchanged.