Investigate best sampler argument
Closed this issue · 0 comments
kailuowang commented
so for now, the thing to do is just to manually parameterize, ie with Walkers as you did before, or if you want, with something like
ap.sample(HMC(5), 1000, 1000)
.... you can do some diagnostic plots etc
like here: https://github.com/stripe/rainier/blob/develop/rainier-example/src/main/scala/com/stripe/rainier/example/DLM.scala#L76
from https://gitter.im/com_stripe_rainier/Lobby?at=5bb3d6c4600c5f642343d15c
https://www.statlect.com/fundamentals-of-statistics/Markov-Chain-Monte-Carlo-diagnostics