and then do the same thing but with different sampler settings and give the inference object a new name
e.g.
mSampled_slice=model0.sample( ...)
then look at the arviz summary it doesn't update the stats and just reiterates whatever inference object was already tied to model0. IT works fine if I define a new model (model1 that is identical to model0) and sample that.
This should be fixed because one might be misled to think that their sampler settings didn't affect anything .
Describe the bug
when I do something like
mSampled=model0.sample( ...)
and then do the same thing but with different sampler settings and give the inference object a new name e.g.
mSampled_slice=model0.sample( ...)
then look at the arviz summary it doesn't update the stats and just reiterates whatever inference object was already tied to model0. IT works fine if I define a new model (model1 that is identical to model0) and sample that.
This should be fixed because one might be misled to think that their sampler settings didn't affect anything .
HSSM version
0.2.2