Two-stage Bayesian regression enforcing a fixed distribution (not Just Hierarchical regression)

[ricardoV94]

@Ray_Kruger

How would you @ricardoV94 @junpenglao suggest inputting a fixed distribution in PyMC? Surely two-stage/sequential regression should be possible?

You would need to assign a custom sampler to the stochastic variable so that the nuts update happens separately from the random draw. Similar to how discrete variables are sampled separately from continuous ones in PyMC