It's interesting to consider what would be involved in getting lazyppl's functionality into monad-bayes. The reason to think this might be quite straightforward is that monad-bayes is especially built around the idea of having an abstract probability monad which is interpreted in a variety of ways. So we could potentially add Prob and Meas from LazyPPL as interpretations.
In particular, Prob in LazyPPL serves a similar role to SamplerIO in monad-bayes, except it is lazy. Meas is just Weighted Prob, so we'd actually get that for free.
It's interesting to consider what would be involved in getting lazyppl's functionality into monad-bayes. The reason to think this might be quite straightforward is that monad-bayes is especially built around the idea of having an abstract probability monad which is interpreted in a variety of ways. So we could potentially add
Prob
andMeas
from LazyPPL as interpretations.In particular,
Prob
in LazyPPL serves a similar role toSamplerIO
in monad-bayes, except it is lazy.Meas
is justWeighted Prob
, so we'd actually get that for free.