Closed THargreaves closed 2 months ago
Ah just need to wait for the release corresponding to this PR to be available.
The current Turing version is forcing Advanced Ps 0.5.6 which doesn't support the new interface.
The current Turing version is forcing Advanced Ps 0.5.6 which doesn't support the new interface.
There is a new release.
Inspired by this gist, this is an implementation of particle-marginal MH using Turing.jl.
Results are compared to the ground truth from a Kalman filter and the posterior means match.
I haven't done any adaption for the MH step and I'm not exactly sure how to, but it seems to work fine with default settings.