Skip to content
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.

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

add multilabel_probabilities #15

Conversation

KimBue
Copy link

@KimBue KimBue commented Jul 6, 2022

Signed-off-by: Kim Buergl

@KimBue KimBue marked this pull request as draft July 6, 2022 13:59
@chschroeder chschroeder marked this pull request as ready for review July 6, 2022 14:16
@KimBue KimBue force-pushed the 14-mulitlabel-clfpredictreturn_proba=true-only-returns-probabilities-for-labels-over-the-threshold branch from d352364 to 2cf55e7 Compare July 6, 2022 14:30
Copy link
Contributor

@chschroeder chschroeder left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, for your effort!

This is going into the right direction. I added some suggestions and found one place where I think some logic is missing.

Otherwise all the classifiers must be adapted (and tested). I would suggest the use the kwarg of the name sparse_proba here (with a default of True).

small_text/utils/classification.py Outdated Show resolved Hide resolved
small_text/utils/classification.py Outdated Show resolved Hide resolved
small_text/classifiers/classification.py Outdated Show resolved Hide resolved
@KimBue KimBue force-pushed the 14-mulitlabel-clfpredictreturn_proba=true-only-returns-probabilities-for-labels-over-the-threshold branch from 2cf55e7 to 147e626 Compare July 8, 2022 09:47
@chschroeder chschroeder marked this pull request as draft July 27, 2022 15:00
@chschroeder
Copy link
Contributor

Closing this after consultation with @KimBue. This will be continued but at another time.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Mulitlabel: Clf.predict(return_proba=True) only returns probabilities for labels over the threshold
2 participants