Jesse Grabowski

It just appears to not work? Can you confirm that `sigma_log__` is in the `posterior` group of the idata if you run this code snippet? ```py import nutpie as ntp...

But the unconstrained value is `sigma_log__`?

Go into `tests/distributions/test_multivariate.py::TestICAR::test_icar_rng_fn` and update the test to no longer expect an error. Check the other `test_rng_fn` tests for how to validate the output. It looks like you also have...

Did you check the PR that Ricardo linked? This would be the preferred way to go.

You would convert the entire distribution into a `SymbolicMVNormalUsedInternally` subclass, like how `MvStudentTRV` was changed in the linked PR (check [here](https://github.com/pymc-devs/pymc/pull/7685/files#diff-ba7699f4e550fe6433c577626dd433a017c2fad177dcbfc83ea8268fe4322761R385) specifically). Then no rng_fn is required at all. The...

You can push the code you have so far and we can look at it here

@MarcoGorelli Nice to meet you today :) I'm going to take a look at this and try to re-aliven the effort, we can maybe do a pair coding in a...

You can definitely take it, but I want to wait for input from other devs before doing anything.

I added a solution to the bug, but I think it might be over-engineered. Basically the problem is that once we're inside the sampling loop, we don't have information about...

The update to the displayed statistics is done [here](https://github.com/pymc-devs/pymc/blob/5db3779b11a1e1d663447a3a78a6f1213b740409/pymc/util.py#L886). The progress manager has access to the return from `iter_sample` e.g. [here](https://github.com/pymc-devs/pymc/blob/5db3779b11a1e1d663447a3a78a6f1213b740409/pymc/sampling/mcmc.py#L1235). This is the current MCMC point (not helpful) and...