Repository navigation
Conversation
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
MAPE zero handling selects near-zero targets, then uses their index labels to drop or replace rows. With duplicate labels,
dropalso removes unrelated observations sharing those labels. On pandas 2.3,replaceraisesInvalidIndexErrorinstead of producing the report.For targets
[0, 1, 2, 0, 4, 8], predictions[2, 2, 1, 3, 6, 8], and index[7, 7, 8, 8, 9, 9], dropping the two near-zero target rows should produce MAPE 50%. The current code drops four rows and reports 25%.Change
Keep the boolean near-zero mask and use it directly for both row filtering and replacement. This preserves the individual observations even when their index labels are repeated, including MultiIndex labels.
The zero threshold, replacement value, denominator formula, and
near_zero_valuescounter are unchanged. In particular, this is separate from #1933 (the near-zero row counter) and #1986 (negative-target denominators).Validation
nonehandling.Report([MAPE(...)])calls verify that changing only row labels does not change the reported metric.git diff --checkpass.The full repository test suite was not run. This change corrects selection of evaluation rows; it does not claim improved model prediction quality.