You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I remember that at some point, we needed to enclose the attribute in backticks to not get an error about undefined references. Then at some point, we didn't need to do that any more. I just upgraded us to numpydoc master, and now I have to add the backticks in a gain.
Were there any intentional changes in that? It's also possible that we hacked this somewhere ourselves, but I can't find the relevant change in the code :(
The text was updated successfully, but these errors were encountered:
I think this is because attribute descriptions used to be generated by SphinxDocString._str_param_list but now they are generated by SphinxDocStr._str_member_list. I'm not entirely sure what the difference is.
This is more of a question.
In scikit-learn we name out attributes with trailing underscores like
I remember that at some point, we needed to enclose the attribute in backticks to not get an error about undefined references. Then at some point, we didn't need to do that any more. I just upgraded us to numpydoc master, and now I have to add the backticks in a gain.
Were there any intentional changes in that? It's also possible that we hacked this somewhere ourselves, but I can't find the relevant change in the code :(
The text was updated successfully, but these errors were encountered: