Fix McNemar ZeroDivisionError when models have no discordant pairs - #795
Open
tonycoder-hub wants to merge 1 commit into
Open
Fix McNemar ZeroDivisionError when models have no discordant pairs#795tonycoder-hub wants to merge 1 commit into
tonycoder-hub wants to merge 1 commit into
Conversation
McNemar's statistic divides by the number of discordant pairs, so comparing two models that are correct and incorrect on exactly the same examples raised a ZeroDivisionError. Report a statistic of 0.0 (and therefore a p value of 1.0) instead, since there is no evidence of a difference in that case. Co-authored-by: Tony Coder <407243179@qq.com>
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
McNemar's statistic is
|b - c|² / (b + c). When the two models being compared produce no discordant pairs (b + c == 0) — comparing a model to itself, or two models that agree on every example — this raisedZeroDivisionError.The comparison now reports
stat = 0.0in that case, which yieldsp = 1.0: there is no evidence of a difference.Tests
python3 -m pytest tests/test_metric_common.py -k "comparison or mcnemar" -q→ 4 passed, 3 skipped.New
test_mcnemar_without_discordant_pairsfails on unmodified source.