“if you express the problem in a more structured way...you can make the sampling scale to large problems (thousands of unknown parameters). For Python there’s #PyMC3 and #PyStan”... #Bayesian twitter: any examples in the wild with >1000 params? Is that computationally possible?https://twitter.com/fulhack/status/1049823636226490368 …
-
-
I've started looking around for this for the arviz project - we plan to implement the same models using multiple different (python) inference libraries, but can ping you once we start pushing code there.
-
I've got a model with 60k parameters that seems to sample fine (most of the parameters could probably be marginalized out though). One item on my things-to-blog-about list. Would be fun to try arviz on that one :-)
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.