Luciano Paz
Luciano Paz
> Hi, Is this the volatile logic in question? [L1698 in sampling.py](https://github.com/pymc-devs/pymc/blob/1417785c6eadb51591409c013a35ae31686031c5/pymc/sampling.py#L1698) > > Looking at the source and being a complete newbie when it comes to development and advanced...
> ```python > group_idx, _ = pd.factorize(data.index) > > with pm.Model() as model: > group_map = pm.Data('group_idx', group_idx, mutable=True) > # non-centered intercept prior > mu = intercept[group_map] > #...
To handle the mixed in and out of sample indexes, we would mainly have to rewrite the `intercept_rv`. The super simplified and pseudocodish way of doing this would look something...
No shame at all @jessegrabowski! I think that the whole point of open source development is to have engaged people, like yourself, speak up and iterate as a community. When...
> Would a warning on `pm.sample_posterior_predictive` that says something like "The shapes of the following random variables have changed, and will be sampled using their prior distributions: ..." be too...
@ricardoV94, just to give some context, this was introduced [here](https://github.com/pymc-devs/pymc3/pull/3300) and it was intended to solve problems that could happen when a value they should be a float was an...
I can only think of one computational disadvantage. If you have a discrete random variable, you should be able to use it to index into an array. Imagine something like...
Are you sure your mingw is the 64bit version? If you installed it from the mingw website, it is very likely that it will be 32bit. Msys2 has 64bit versions...
Yes, you installed the 32 bit version. I’m not an expert on this, but I know that msys2 ships a 64 bit version of mingw, whereas the mingw site only...
What would leave future or need to be able to work? From a consistency point of view, I think that the consistent thing to do is to collapse the logp...