DOC improve the UG documentation of the classification report #868
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
…mbalanced` output contains Especially point out that the Weighted Average Recall that `classification_report_imbalanced` displays is also better known as the Accuracy! Maybe this should also be a change in the API docs for `classification_report_imbalanced` but the "Edit" link there didn't work.
|
Hi, the full title for this proposed change got truncated (sorry if I made it too long). It was supposed to be: |
glemaitre
changed the title
glemaitre
approved these changes
Dec 7, 2021
There was a problem hiding this comment.
Thanks @DavidRosen. LGTM. I merge master into your branch and make minor changes.
I will merge once the CIs are happy.
Codecov Report
@@ Coverage Diff @@
## master #868 +/- ##
==========================================
- Coverage 97.18% 94.36% -2.83%
==========================================
Files 96 96
Lines 6253 6155 -98
Branches 712 584 -128
==========================================
- Hits 6077 5808 -269
- Misses 104 283 +179
+ Partials 72 64 -8
Continue to review full report at Codecov.
|
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Outdated suggestions cannot be applied.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.
…mbalanced` output contains
Update metrics.rst -- say what bottom row of
classification_report_imbalancedoutput contains...and especially point out that the Weighted Average Recall as displayed by
classification_report_imbalancedis also better known as the Accuracy! Maybe this should also be a change in the API docs forclassification_report_imbalancedbut the "Edit" link there didn't work. BTW the Accuracy and Wtd Avg Recall are also the same as the Micro Avg Recall but I'm not sure if that's important so I didn't mention it in the update.Proof that Weighted Average Recall == Accuracy:
Let Ntot be total number of instances. Let iclass be the class index.
Note that recall(iclass) * support(iclass) = number of instances in iclass classified correctly.
Wtd Avg Recall = sum over all iclass (recall(iclass) * proportion(iclass) ) {by definition of Wtd Avg Recall}
= sum over all iclass (recall(iclass) * support(iclass) / Ntot )
= sum over all iclass (recall(iclass) * support(iclass) ) / Ntot
= sum over all iclass (number of instances in iclass classified correctly) / Ntot {from Note above}
= total number of instances classified correctly / Ntot
= Accuracy {by definition of Accuracy)
Reference Issue
What does this implement/fix? Explain your changes.
Any other comments?