Colin Carroll
Colin Carroll
Sounds great! Maybe worth updating https://github.com/arviz-devs/arviz/wiki/ArviZ-2021-roadmap to keep track of your thoughts? I like the idea of arviz's non-plot code providing a sufficient API to (easily) reproduce the plots: seems...
https://github.com/stan-dev/stan/wiki/Prior-Choice-Recommendations
(assigning myself to link that in the appropriate spot)
Maybe wrapping in a `try/except ValueError` right [here](https://github.com/arviz-devs/arviz/blob/main/arviz/__init__.py#L318)? It doesn't "commute" with import order, but I'll pop over to the `colorcet` side and see if maybe that should just be...
It is the same colorscale, so it makes sense from that perspective.
Sounds like they are suggesting something similar to you, @OriolAbril ! https://github.com/holoviz/colorcet/issues/96#issuecomment-1276576516
This could push https://github.com/arviz-devs/arviz/pull/2086 over the line, and unblock both this feature and https://github.com/arviz-devs/arviz/pull/2040 Does this feature actually need two functions for computing posterior density and a constraining bijector, or...
That's a good question! I haven't doublechecked, but I think you get sum(`2 ** tree_depth`) logp and grad(logp) evaluations. This is an approximation, I think because: 1. (most of the...
With, e.g., blackjax, don't we already lose the instrumentation inside the actual sampler (other than what is provided?) Seems like if the pymc sampler implemented a gradient or logp counter...
(I tagged @ericmjl even though he's busy in case he has wisdom to share here, but no pressure!)