How to model historical data in the right distribution

Questions version agnostic
2

I had this issue as well. Check out this thread What is the best way to estimate theta and psi for ZeroInflatedPoisson? - Questions / version agnostic - PyMC Discourse