Add accepted arg to AdvancedMH.Transition calls

[sunxd3]

Reflect TuringLang/AdvancedMH.jl#88

[devmotion]

As I wrote in TuringLang/AdvancedMH.jl#88 this approach can easily lead to misleading/incorrect acceptance rates?

[sunxd3]

The field accepted is added, but seemingly not used anywhere right now. I can see it causing issue particularly with thinning, but maybe we can always return the acceptance rate of the un-thin-ed chain? I am certain here

[sunxd3]

Anyhow, this PR is to just patch the interface so that it won't error.

[codecov]

Codecov Report

Attention: 3 lines in your changes are missing coverage. Please review.

Comparison is base (990eeb2) 0.00% compared to head (6d2f291) 0.00%.
Report is 1 commits behind head on master.

Files	Patch %	Lines
src/mcmc/mh.jl	0.00%	2 Missing ⚠️
src/mcmc/emcee.jl	0.00%	1 Missing ⚠️

Additional details and impacted files

@@          Coverage Diff           @@
##           master   #2172   +/-   ##
======================================
  Coverage    0.00%   0.00%           
======================================
  Files          22      22           
  Lines        1385    1385           
======================================
  Misses       1385    1385           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[yebai]

As I wrote in TuringLang/AdvancedMH.jl#88 this approach can easily lead to misleading/incorrect acceptance rates?

Can we print a warning message indicating that the acceptance rate might be inaccurate if thinning or discarding warmup is applied?

[yebai]

Better to use a more informative file name here.

[sunxd3]

Oops, sorry, this file should not be here, it was a local script, I'll remove it