Currently the code breaks somewhere if the user passes a dict where some distribution parameters are missing, e.g. do_scale = {1 : True}. Now we raise a warning and fill the default.
On the same go, I'd like to make forget and estimation_kwargs dictionaries that can be individual for each distribution parameter. Potentially also method but that includes a bigger refactoring.
Currently the code breaks somewhere if the user passes a
dict
where some distribution parameters are missing, e.g.do_scale = {1 : True}
. Now we raise a warning and fill the default.On the same go, I'd like to make
forget
andestimation_kwargs
dictionaries that can be individual for each distribution parameter. Potentially alsomethod
but that includes a bigger refactoring.