I want to implement two-stage Bayesian regression. Where one models posterior feeds into the second model (i.e separate regression). It appears that PyMC only allows free random variables that are to be inferred or observed data and cannot take “fixed distributions” to be sampled as input. What I …

In any case, @Ray_Kruger this is a proof of concept you can take a look: Say you have model at first level as: mu0, sigma0 = 10., 1.5 y0 = np.random.normal(mu0, sigma0, size=1000) with pm.Model() as m: mu = pm.Normal('mu', 0, 100) sigma = pm.HalfNormal('sigma', 10) y = pm.Normal('y', …

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

Questions v5

junpenglao July 29, 2022, 5:48am 2

I think you are on the right track, but probably it is possible to use the RandomVariable from aesara directly. Maybe @ricardoV94 could advice a bit more?

Topic		Replies	Views
Custom sampling step Questions	36	2637	November 8, 2022
Hierarchical linear model with two parent distributions Questions	16	1338	September 15, 2018
Getting observations of distribution variables whilst sampling Questions from_github	2	1008	June 15, 2017
PyMC3-Sampling not taking place for the entire range of values between the given bounds of a Uniform prior Questions	7	1606	January 14, 2018
DensityDist conditional random() method Questions development	7	1093	July 5, 2019

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

Related topics