New feature: random seed for python models

[eozd]

Thank you for providing this package!

I was wondering whether it is possible to specify a random seed in the python version of hBayesDM. In the current master version I haven't found a way of passing the seed to the stan model. I guess the easiest way of adding this functionality would be to pass the seed parameter to

hBayesDM/Python/hbayesdm/base.py

Line 708 in 2781155

return sm.vb(data=data_dict,

and to

hBayesDM/Python/hbayesdm/base.py

Line 712 in 2781155

return sm.sampling(data=data_dict,

What do you think about this feature? Hoping to hear back from you soon.

[JaeyeongYang]

Sorry for the late reply. That sounds good! I will add this option in the next version release. Thanks for your suggestion. :)

[deschsimon]

I was wondering about this very same feature in the R version.
There I guess a seed value was best passed to stan at

hBayesDM/R/R/hBayesDM_model.R

Line 473 in d11ddb4

fit <- rstan::vb(object = stanmodel_arg,

, and

hBayesDM/R/R/hBayesDM_model.R

Line 478 in d11ddb4

fit <- rstan::sampling(object = stanmodel_arg,

, respectively.

Could this be easiliy implemented?