8000 typo in AUC macro documentation model_evalution · Issue #18515 · scikit-learn/scikit-learn · GitHub
[go: up one dir, main page]
Skip to content

typo in AUC macro documentation model_evalution #18515

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
reshamas opened this issue Oct 2, 2020 · 6 comments · Fixed by #18559
Closed

typo in AUC macro documentation model_evalution #18515

reshamas opened this issue Oct 2, 2020 · 6 comments · Fixed by #18559
Labels
Documentation

Comments

@reshamas
Copy link
Member
reshamas commented Oct 2, 2020

Describe the issue linked to the documentation

(cwjbrian50)

I am a good user of scikit-learn. First of all, I am grateful to
scikit-learn for providing a good service. The reason I write an email is
to report some typos.

While studying about AUC, I think I found some typos in API documentation.
As it is written in "
https://scikit-learn.org/stable/modules/model_evaluation.html#roc-metrics"
One-vs-one Algorithm, the multiclass macro AUC metric was defined in the
reference [HT2001]. But there is a double difference in macro ovo AUC
between the reference and documentation. Furthermore, under the macro AUC,
there is weighted AUC which is defined in the reference [FC2009] as it is
written in documentation. But there is no same metric in the documentation
in the reference however there is a similar one, AU1P. After reviewing the
code which includes the definition of roc_auc_score, I notice that it is
different from both of the expressions in the documentation and the
reference. Additionally, my scikit-learn version is 0.23.1. I hope this
part will be fixed well. Thank you for reading my email.

Suggest a potential alternative/fix

the best way to get this fixed is to send us a PR updating this file:
https://github.com/scikit-learn/scikit-learn/blob/master/doc/modules/model_evaluation.rst
thanks for your help
@RamyaNP
Copy link
Contributor
RamyaNP commented Oct 7, 2020

Hi, I'm new to contributing to open source. i wanted to try and help fix this issue.
If no one else is working on this, may I?

@reshamas
Copy link
Member Author
reshamas commented Oct 7, 2020

Hi, I'm new to contributing to open source. i wanted to try and help fix this issue.
If no one else is working on this, may I?

@RamyaNP Sure, go for it.

@RamyaNP
Copy link
Contributor
RamyaNP commented Oct 7, 2020 •
edited
Loading

@reshamas
I went over the documentation page and the reference page that had been mentioned in the issue. But I am unclear on what the following lines mean. Could I get some elaboration on this?

But there is a double difference in macro ovo AUC between the reference and documentation.

@reshamas
Copy link
Member Author
reshamas commented Oct 7, 2020

I copied this issue over from the mailing list. I would have to examine it more closely to work on it, and unfortunately don't have the time to look more closely.

@thomasjpfan
Copy link
Member

The two here should be a one:

Screen Shot 2020-10-07 at 9 09 37 AM

Same for the weighted case.

@RamyaNP RamyaNP mentioned this issue Oct 7, 2020
Merged
@RamyaNP
Copy link
Contributor
RamyaNP commented Oct 7, 2020

Thank you @thomasjpfan
I've made changes accordingly, please tell me if what I have done works.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
Documentation
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

DOC fix multiclass AUC formulas in user guide RamyaNP/scikit-learn
3 participants
@reshamas @thomasjpfan @RamyaNP
0