Closed danielturek closed 11 months ago
I'm fine with anyone merging this any time, I have confidence in the (updated) RW_multinomial
sampler. Presumable if I just leave this open, @paciorek will merge it at some point.
Merged manually to deal with conflicts.
Fixes dubious results in
RW_multinomial
sampler, by correcting the reverse proposal probability in the MH acceptance probability calculation.This addresses NCT issue # 486.