You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
MH is implemented in a way where even if it receives a AdvancedMH.RandomWalkProposal it will still hit the "general" propose!! that assumes the proposal is a NamedTuple, i.e.
* fixed dispatch issue of `propose!!` (ref #2180)
* added test for `filldist` proposal for MH
* bump patch version
* added `demo_dot_assume_observe_index` to models that are allowed to
fail the MLE test, as it should be
* fixed test
* try comparing median instead of mean to avoid outliers completely
ruining the estimate
* try with initial params
* upped the number of samples used for `filldist` proposal test
MH
is implemented in a way where even if it receives aAdvancedMH.RandomWalkProposal
it will still hit the "general"propose!!
that assumes the proposal is aNamedTuple
, i.e.Turing.jl/src/mcmc/mh.jl
Lines 377 to 383 in 3a315ce
instead of the one that is hit if we're a subtype of
MvNormal
Turing.jl/src/mcmc/mh.jl
Lines 410 to 417 in 3a315ce
For example, if I want to use a
filldist(..., n)
as a proposal instead ofproduct_distribution(Fill(...), n)
, stuff breaks:IIUC we should just relax
Turing.jl/src/mcmc/mh.jl
Lines 410 to 417 in 3a315ce
to be any
AdvancedMH.RandomWalkProposal
🤷The text was updated successfully, but these errors were encountered: