-
-
Enhance other (Environment, Metropolis Hastings, inverse problem) constructors with options setting flexibility similar to GPMSA, so one can default construct, set via C++ API, or parse from input fil…
-
-
La idea de este issue es que dejemos un enlace, nota, imagen, comentario, etc. a recursos que nos encontramos por ahi y pensamos que pueden ser utiles para desarrollar este curso.
-
Potential alternative to MH, supposed to be better.
A variant is the "NUTS" algorithm, which is implemented in the package "stan"
Should look into at some point.
-
I've been having a look at the MCMC code in Breeze, and it seems like it could do with some tidying up. There are currently two separate implementations, which is confusing, neither of them are well-d…
-
Last night, @cameronfreer stumbled across a bug in the tutorial notebook:
```
(define [_ mh-tr _]
(infer :procedure single-site-metropolis-hastings-step
:inputs [circus-brothers [] i…
-
I'm running into this for my custom grammar. (PS: Trying to reproduce list processing task of DreamCoder).
Any tips on how to approach grammar definition that avoids this?
-
I usually code my MCMC algorithms (mostly gibbs in BNP) in C++, but this time I thought I might switch to tfp for a change.
Is there anyone using tfp for (blocked) gibbs sampling / some resources onl…
-
```
bw = Exponential(0.1)
y = Exponential([1. / bw, 1. / bw])
ys = tf.stack([Exponential(1. / bw), Exponential(1. / bw)])
```
Both `y` and `ys` have the same distribution. Metropolis-Hastings…