search

pulearn / pulearn

Positive-unlabeled learning with Python.
https://pulearn.github.io/pulearn/
BSD 3-Clause "New" or "Revised" License
218 stars 33 forks source link

feat: implement metrics #19

Closed kjappelbaum closed 2 years ago

kjappelbaum commented 2 years ago

partially addresses #12

also updates documentation to address #8

shaypal5 commented 2 years ago

Hey @kjappelbaum

This looks amazing!

Please rebase over my latest commit to master - I've moved testing to Github Actions.

Then, we should be able to see run results here. If we won't, we'll close this one and you'll open a new PR.

Also, please fix the single issue CodeFactor found (an extra space after the = in line 91 of metrics.py:

image
codecov-commenter commented 2 years ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 100.00%. Comparing base (8ca1ca1) to head (707f215). Report is 31 commits behind head on master.

Additional details and impacted files ```diff @@ Coverage Diff @@ ## master #19 +/- ## ========================================= Coverage 100.00% 100.00% ========================================= Files 2 3 +1 Lines 322 334 +12 Branches 50 54 +4 ========================================= + Hits 322 334 +12 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

🚨 Try these New Features:

shaypal5 commented 2 years ago

Hey @kjappelbaum

Few issues here:

  1. Commit-wise, please rebase your whole PR into a single commit, that lies AFTER the current head of master.

  2. Note I've ammended the latest commit (to hopefully enable the linting job to leave annotations on PRs from forks), so you will have to rebase carefully. I suggest you simply realign your master with my master (will probably require git reset --hard origin/master on your master; just make sure all your work is also backed-up outside the repo's folder), then checkout a new metrics-new branch and just using a simple file system copy-paste to override all the latest files with your changed files (and ONLY with your changed files). Then just stage everything into a single commit with the combined PR description you have here as the commit message.

  3. Test coverage is kept to a 100% on pulearn but your codes bring it down to 99.4% (because of a 83.33% coverage on metrics.py). Please bring it back up to a 100%.

  4. You have many linting errors. Either run flake8 yourself, or checkout the linting anotations here (assuming my latest commit will indeed fix the issue).

Just a little more effort! Thank you for contributing! <3

kjappelbaum commented 2 years ago

Hi - sorry, was on mobile and this caused a somewhat messy commit history.