Closed rzshrote closed 2 years ago
Issue was with how function aliases weren't accessing the global random number generator for numpy
. Fixed by eliminating wrapper class to handle seeding and spawning of new generators. The new implementation is a simpler and more robust wrapper module (no wrapper classes within the module though).
Setting the seed does not result in deterministic results. I am not sure where this is coming from. This error appears to be directly related to the seed function.
Sample output: