Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
Numpy conventions #226
I thought about splitting it to 2 parts:
@Nurdok , your opinion?
Jan 20, 2017
1 check passed
@farmersez - Thank you very much!
Having some docs for this would be great!
This was referenced
Jan 22, 2017
referenced this pull request
Jan 22, 2017
This looks interesting, @farmersez. However, in practice what we need in scikit-learn and other numpydoc users are checks for parameter listings:
Sorry for the really really really late response
@cdeil, I ran it on the numpy example file (after you caught the D410 bug
And to follow up @jnothman 's remarks:
When I started the numpy validation feature I tried to do all that - pass function parameters to the docstring validation code and check its structure, but in my opinion it got too entangled and was not worth the trouble at the time. I do intend (and/or encourage you) to add actual content validation to sections.
Having said that, the validation I'm talking about is strictly for numpy. If your project has a different documentation style for sections perhaps we should consider a way to allow external validation code running within pydocstyle, though I don't believe that's a thing worth doing. @Nurdok? entry points maybe?