Update docstrings for better rendering on website and numpydoc compliance

[OriolAbril]

Overview

We have to review and update all the docstrings of objects part of the public API (that is, that appear on the API reference section of the website). Why?

There are two main reasons:

• Some functions are not documented correctly. i.e. pymc.Uniform didn't document that lower and upper bounds are optional and default to 0 and 1 respectively
• Not following numpydoc results in non-ideal rendering. i.e. if there are no spaces around the colon that separates parameter name and type numpydoc+sphinx can't render it correctly and the type ends up being shown as bold text instead of being a link to the documentation of that object (which is particularly useful for functions taking or returning external objects such as ArviZ InferenceData or aesara TensorVariable).

Current view of the parameter section of pymc.Uniform

Before doing the changes described below

Details

A detailed tutorial on how to go about these updates is available at the PyMC-Data Umbrella sprint website (the page also has a video recorded by @reshamas showing an example PR).

How to choose what page to work on?

Look for pages that don't have hyperlinks in all their parameters. In case of doubt, don't hesitate to ask in this issue for pages to work on.

[ravichikkam]

I am working on the docstrings for weibull distribution.

EDIT (by @OriolAbril ): I moved the updated list to the initial comment so its easier to find for new people getting to this issue

[jmloyola]

Working on pm.Data.

[cuchoi]

Is this where the ConstantData and MutableData docs should appear? https://docs.pymc.io/en/v3/api/data.html. I can't find them.

Edit: There in v4 here https://docs.pymc.io/en/latest/api/data.html

[chritter]

Working on pymc.Exponential

[cuchoi]

Going to work on pymc.HyperGeometric

[sagartomar]

Working on pymc.Normal

[lalitmusmade]

Working on pymc.Flat

[GallowayJ]

Working on pymc.GaussianRandomWalk

[areding]

I'll start on pymc.Gamma

[areding]

Starting pymc.InverseGamma.

[sagartomar]

Starting with pymc.Beta

[lalitmusmade]

Working on pymc.HalfFlat

[bdatko]

@jlindbloom and I are working on Other utils

[pibieta]

I am working on the class docstrings inside the multivariate module.

[vitaliset]

Me and @pibieta are working on the rest of the classes docstrings inside the multivariate module. #DataUmbrellaPyMCSprint Cc: @reshamas

[daniel-saunders-phil]

working on docstring for pm.simulator

[rowan-schaefer]

I can take smc

[daniel-saunders-phil]

I can take smc

maybe we could work together on it! I've been working with a little bit this week and have a number of ideas. Do you know a good technique to co-write a PR?

[rowan-schaefer]

I'll message you!

[daniel-saunders-phil]

I'll take pm.AR

[rowan-schaefer]

Currently working on pymc.variational.inference!

[hyosubkim]

I will work on pymc.DiscreteWeibull

[apalermo01]

I'm getting pymc.model.core.Model

[HangenYuu]

I will work on pymc.step_methods.BinaryGibbsMetropolis at this link: https://www.pymc.io/projects/docs/en/stable/api/generated/pymc.step_methods.BinaryGibbsMetropolis.html