Repository navigation
Doc build warning from io.table.Column - #280
Conversation
|
This would explain the problem I'm having in #279. Good to know it wasn't my fault :) |
|
I generally lead to not documenting things from superclasses, but having a nice obvious link to the documentation for the superclass. But I know people who are less comfortable with OO generally prefer to see everything they can do with an object. |
|
@mdboom - I can see it either way, too. For now, I'll try to get it to strip the superclass attributes, and include a link somewhere to the superclass, as that should be pretty easy to do quickly. We can re-visit the question later if people don't like it. |
|
I think that in this case the ndarray docstring is so extensive, it makes it very hard to find out what you can do specifically with the Column class. Better to leave all the ndarray stuff out and just leave in a link to the Numpy docs. |
|
These commits now implement as discussed here - class doc pages show links to the base class, but base class attributes and methods are not displayed. Note that this is only true for classes documented using I will merge this tomorrow before release if there are no objections. |
Fix build warning in io.table.Column by supressing base class attribute documentation
Fix build warning in io.table.Column by supressing base class attribute documentation
Fix build warning in io.table.Column by supressing base class attribute documentation
Improve the description of pyspeckit
After #268 was merged, there's now a sphinx warning about the
Columnclass. The root of the problem is that autosummary is trying to document all the attributes ofColumnnow, andColumnis a subclass ofnumpy.ndarray... and some of the ndarray attributes don't play nice with our doc standard, apparently.The larger issue here is how/whether to document attributes from superclasses. I think at least for now we should just supress those attributes, but it's an issue to consider further.