Luciano Paz

MarginalSparse shouldn't use DensityDist internally

2

## Description of your problem The `MarginalSparse` Gaussian Process uses `DensityDist` internally to compute the `marginal_likelihood` ([here](https://github.com/pymc-devs/pymc3/blob/main/pymc/gp/gp.py#L733-L756)). We should refactor that code to use a proper RandomVariable instead ## Versions...

Refactor step methods to use their own random stream

2

## Description of your problem This issue is a spin off from a discussion in #5787. At the moment of that PR, our step methods rely on the global numpy...

Add a way to plot credible intervals with plot_ppc

6

## Tell us about it `plot_ppc` is very nice, you can control the `kind` of plot and also the number of posterior predictive lines to draw with `num_pp_samples`. I would...

Write a notebook on manual model vectorization

1

After #193, I never got around to writing a proper usage notebook, so I'm opening this issue as a reminder for when I have some time to do it. This...

Class level dimension and parameter information

1

I opened this issue just to keep track of [this pymc3 discussion](https://github.com/pymc-devs/pymc3/issues/3984). Whatever design choice we end up taking there, should also be addressed here in pymc4.

This closes #264. However, it highlights that some of the current docstrings are broken, so we will have to fix them before being able to merge. Also, we need to...

The distributions transforms name is misleading

2

### Background Following some discussions that took place on our discourse channel, over at pymc3 and in particular in #246, it came to our attention that the `transform` argument of...

This closes #142 and it should only be merged once #253, #254, #255, #256, #257, #259, #260, #261, #262, and #263 are fixed.

We should test that our docstrings can be rendered into a simple apidoc without warnings using sphinx.

At the moment, running `pydocstyle --convention=numpy pymc4/distributions` will raise many errors. In particular in the continuous, discrete and multivariate modules. We need to fix that in order to address #142.