ENH: add dtype option to numpy.lib.function_base.cov and corrcoef#17456
ENH: add dtype option to numpy.lib.function_base.cov and corrcoef#17456rossbar merged 15 commits intonumpy:masterfrom
Conversation
Co-authored-by: Eric Wieser <wieser.eric@gmail.com>
Co-authored-by: Eric Wieser <wieser.eric@gmail.com>
Co-authored-by: Eric Wieser <wieser.eric@gmail.com>
eric-wieser
left a comment
There was a problem hiding this comment.
This generally looks good, thanks for the contribution. Two more things:
- The
dtypeargument needs documenting, and should include.. versionadded:: 1.20.0 - We should maybe add a release note in the
upcoming_changesfolder.
|
Sure, I can certainly do that. I left it out initially to see whether there was interest in actually merging the PR. Thank you for reviewing so quickly @eric-wieser ! |
eric-wieser
left a comment
There was a problem hiding this comment.
This probably doesn't need to hit the mailing list, but I'll leave it for another maintainer to give a second opinion on. Thanks for the rapid iteration.
Co-authored-by: Eric Wieser <wieser.eric@gmail.com>
|
Happy to help! Its a bit silly maybe but if its no trouble, could you add the label "hacktoberfest-accepted" to the PR? They changed the rules to prevent spam. |
|
@lschwetlick, there's no need for me to do that according to your link, my approval above is enough. What does matter is if the repo has a |
|
Yup, I get it. I mainly liked the gamification aspect as a motivator. Good use of my time anyway :) |
rossbar
left a comment
There was a problem hiding this comment.
LGTM as well, seems like a nice improvement. Thanks @lschwetlick !
dc1b5e5 to
3738fae
Compare
|
I've updated the docstring wording based on @seberg 's comment (and undid my test changes since I was in there). I'll merge once the CI's done - I don't want to let wording block this nice new feature (thanks @lschwetlick !) We can always revisit the docs later to improve the wording or add more explanation if needed. |
|
Added the hacktoberfest-accepted label, since I think @mattip said that was preferable to the repo topic. Apologies if it's too late to make a difference! |
|
Nice, thank you @eric-wieser ! |
I tried to implement the changes suggested in Issue #17395, by adding a dtype keyword to the cov and corrcoef functions in numpy/lib/function_base.py. I also wrote tests for the changes.
Closes #17395