Seth Axen

All diagnostics are now in MCMCDiagnosticTools.jl. All stats except `kde` are now here in the ArviZStats module (or re-exported by it in the case of PSIS), which will be split...

I wonder actually if this is going the wrong way about this. MCMCChains destructively flattens draws into one large array, and a converter then needs to infer from the variable...

Is there a reason we need access to the sampling and log-density functionality of a VI model? e.g. when we need samples from a posterior represented by a PPL, we...

> @sethaxen can you take a look at this PR before it gets merged? Certainly! I'll do a proper review in the next few days, but from trying it out...

I just had the same error. > The output of `npm dev` is hidden by the command; this is probably throwing an error. You can try going to the `Pollen.FRONTEND_DIR`...

Currently the Infiltrator example is showing the same error in the search bar: https://juliacomputing.github.io/MultiDocumenter.jl/inf/stable/. This is the main blocker to me using this package.

For anyone else running into the same issue, this hack worked for me: https://github.com/TuringLang/turinglang.github.io/pull/46

Do you anticipate doing so would cause any performance regressions? Otherwise, perhaps it would be better to add a new subtype of `AbstractBandedBlockBandedMatrix`.

Might not be related, but even without jit compilation and complex inputs, gradient computation of a function of eigenvectors fails. ```python import numpy as onp import jax.numpy as np from...

Thanks @hawkinsp for the quick response and fix! Looks like repeated eigenvalues is probably the cause of the second bug ```python import numpy as onp import jax.numpy as np from...