[SciPy-dev] scipy doc milestones duplicates and triplicates

Ralf Gommers ralf.gommers@googlemail....
Thu Oct 22 03:16:28 CDT 2009


On Thu, Oct 22, 2009 at 5:55 AM, <josef.pktd@gmail.com> wrote:

> I looked at http://docs.scipy.org/scipy/Milestones/Milestones_11/
>
> and the list is huge, however, many functions are in there either as
> duplicates (those imported into scipy.stats.*) or as triplicates
> (distributions).
>
> Are functions supposed to be listed in the module, e.g.
> scipy.stats.stats, or the package name, scipy.stats?
>

Can you give an example of duplicates here? I see functions like cov and
itemfreq only listed once.

The pages are under scipy.stats.stats but the ones that are imported to
scipy.stats should be listed only there. That's how the NumPy Milestones
work anyway.

>
> Individual distributions are listed as class, and twice as instances
>
> # scipy.stats.distributions.nbinom
> # scipy.stats.distributions.nbinom_gen
> # scipy.stats.nbinom
>
>
This looks like a bug in pydocweb to me, since

>>> sp.stats.nbinom is sp.stats.distributions.nbinom
True

but there are actually separate pages:
http://docs.scipy.org/scipy/docs/scipy.stats.distributions.nbinom/
http://docs.scipy.org/scipy/docs/scipy.stats.nbinom/


None of the docstrings of the individual distributions can actually be
> edited since they are autogenerated (except for the extradoc).
>

They can be edited in the wiki, but you're right that there will be a
problem when trying to merge wiki edits back into trunk. There's a similar
problem in ndimage, where the wiki collides with docstring manipulation in
the code. No solution for that yet, but it should be fixed somehow.

Cheers,
Ralf


> Any idea what the milestones should look like in this case, or is it
> too early to ask?
>
> Josef
> _______________________________________________
> Scipy-dev mailing list
> Scipy-dev@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.scipy.org/pipermail/scipy-dev/attachments/20091022/fd0f878f/attachment.html 


More information about the Scipy-dev mailing list