Has anyone ever calculated posterior predictive p-values in PyMC3?
Nope but sounds like a bad idea 
what is the context of doing this?
They are used occasionally in the psychology literature: https://dx.doi.org/10.1214/aos/1176325622
Has anyone ever calculated posterior predictive p-values in PyMC3?
Nope but sounds like a bad idea what is the context of doing this?
They are used occasionally in the psychology literature: https://dx.doi.org/10.1214/aos/1176325622