RV-Dependent Observations from Irregularly-Shaped Dataset

Maybe a high order hidden markov model? It might not be trivial in PyMC3 tho…