Alexandre Andorra

Results 82 comments of Alexandre Andorra

Adding the median seems nice, especially to easily spot potentially skewed distribution. I'm not sure about adding MAD though, as the HDI is already in `summary` IIRC, and the latter...

Hi @chrisjsewell, and thanks for your quick answer! I just tried that and unfortunately get the same error as above 😕 Do want me to try something else?

> if you bear with me, I will try to look over the weekend Yes of course, thank you! > If you could perhaps attach/copy here a full notebook from...

Hey @chrisjsewell ! Just checking in to see if my previous comment has enough info for you to reproduce, or if you need anything else? Thanks in advance and take...

Hi @alexlyttle, and thanks for the ping @ericmjl ! I don't think this is on our radar, as PyMC4 is not production ready yet. That being said, the fact that...

Thanks for the first review @ricardoV94 ! I added some tests. How do they look like?

> This would make more sense in multivariate even though we are treating it as a scalar as a hack for the time being. Good point. I moved it to...

There is a weird error on the [ubuntu tests](https://github.com/pymc-devs/pymc/actions/runs/3073048250/jobs/4964929407#step:7:818): ```python > self.step_size = step_scale / (size**0.25) E ZeroDivisionError: float division by zero ``` It's not even coming from ZSN code,...

@lucianopaz @ricardoV94 you can re-review. I'll try to add the test against MvNormal logp this weekend, but other than that this PR should be complete 🤞

Do we have a list of time series distributions already ported to v4 (I'm updating the release notes)