Native torch metrics

[justusschock]

Before submitting

• Was this discussed/approved via a Github issue? (no need for typos and docs improvements)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure to update the docs?
• Did you write any new necessary tests?
• If you made a notable change (that affects users), did you update the CHANGELOG?

What does this PR do?

Starts the implementation of metric functionals.
Fixes #1295
Fixes #1297
Fixes #1296
Fixes #1302

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃

Remaining Todos:

• Modular Interface
• docstrings
• Tests

Regarding the commits: Everything before b193059 should already be on metrics branch, don't know why they are listed here...

[pep8speaks]

Hello @justusschock! Thanks for updating this PR.

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2020-06-12 16:50:54 UTC

[Borda]

seems like there are some additions from already merged in metrics

[justusschock]

@Borda yes, but I don't know why. I've based this branch on metrics

[Borda]

@justusschock resolved the large diff, pls check it...

[justusschock]

There are probably some minor issues with auc calculation, since roc and pr_curve are fine, but aurora and average precision aren't.
Will investigate this further

[Borda]

@williamFalcon why did you close the metrics PR? @justusschock?

[Borda]

@cuent mind to help to finish this PR?

[justusschock]

@cuent wanted to take this over, I think. For now we've been communicating on slack about this :)

[Borda]

cuent wanted to take this over, I think. For now we've been communicating on slack about this :)

cool, let me know if I help you with something around... :]
@cuent mind drop a comment here so I could re-assign it to you...

[Borda]

@justusschock mind turn it for review or rebase it?
as it is 351 commit after master and it has also 91 own commits I think that there will become collisions and sooner to resolve it better :] (also as it is root repo branch @cuent cannot commit directly, just make a PR to this branch so resolving may help)

[cuent]

Yeah I am rebasing and checking some of tests. There are a lot of conflicts :/ I am solving them

[Borda]

Yeah I am rebasing and checking some of tests. There are a lot of conflicts :/ I am solving them

@cuent can you push to this branch?

[cuent]

This PR #1962 should fix tests and rebase

[Borda]

@justusschock so this shall be closed in favour of #1962, right?

[codecov]

Codecov Report

Merging #1488 into master will increase coverage by 1%.
The diff coverage is 94%.

@@           Coverage Diff           @@
##           master   #1488    +/-   ##
=======================================
+ Coverage      86%     87%    +1%     
=======================================
  Files          79      81     +2     
  Lines        4960    5218   +258     
=======================================
+ Hits         4271    4547   +276     
+ Misses        689     671    -18     

[SkafteNicki]

LGTM