Currently, most callers of bayesdb_get_generator (in the soon-to-be-merged 20160624-riastradh-cgpm branch) are talking about generators of a specific population. But there's nothing that forces the generator to be of that implied population. We should add logic to catch the mistake of naming a generator for another population, which requires two different error branches / exceptions: one for nonexistent generator, one for generator of wrong population.
Currently, most callers of
bayesdb_get_generator
(in the soon-to-be-merged 20160624-riastradh-cgpm branch) are talking about generators of a specific population. But there's nothing that forces the generator to be of that implied population. We should add logic to catch the mistake of naming a generator for another population, which requires two different error branches / exceptions: one for nonexistent generator, one for generator of wrong population.