Seth Axen

> It seems UCV and BCV, can be quite noisy compared to "default" except when distribution are normal-like..., so their main use case could be multimodal densities? Did you compared...

Yeah, from a brief survey, it doesn't look like a difficult refactor; but the implementation of the plotting logic for e.g. Plots.jl using those arguments is quite a daunting task,...

Also, would it be better to work on this here or in arviz-stats?

As far as I can tell, given a density $p(x \mid \theta) = f(x, \theta) g(\theta) h(x)$, it would be useful to have the following functions - `logupdf(x) = `...

> Also, maybe this shouldn't be just merged into main? How does this go, @sethaxen? Pathfinder follows a continuous deployment model, so yes we'll merge this directly into main and...

> I'm guessing this makes a bunch of tests fail which rely on the previous behavior? Seems to only be 2 tests: https://github.com/mlcolab/Pathfinder.jl/blob/20dd77e5514c2d5d65a2a5edfb2cf72e1dfb9758/test/singlepath.jl#L34-L35 https://github.com/mlcolab/Pathfinder.jl/blob/20dd77e5514c2d5d65a2a5edfb2cf72e1dfb9758/test/multipath.jl#L38-L39 Not certain if equality check would...

Docs build failure is unrelated, and I'll fix in a separate PR

> The new test would probably have to check that constructing a `default_optimizer` pre and post a pathfinder run results in identical-in-value but different-in-memory objects, if that makes sense. I...

@nsiccha anything I can do to help out with this PR?

> Ah, I've been trying to construct a test that fails for `multipathfinder`, but was unable to. The reason being you already catching the `multipathfinder`+stateful optimizer case, see Ah, yes,...