I’m getting strange behavior with a very simple model with a categorical likelihood. I expect that the issue is my lack of experience with this particular kind of model. Here a minimal example that illustrates the problem: import numpy as np import pymc3 as pm import theano.tensor as tt import mat…

Problem with pm.Categorical

junpenglao November 28, 2017, 9:02pm 4

lol. BTW you might want to use a Dirichlet prior so it generates unit vector p

Making a query for a simple BN made in PYMC3

Topic		Replies	Views
Categorical Prior Distribution Questions	6	1168	March 1, 2019
Strange error with Categorical distribution Questions	3	485	August 15, 2018
Create an array of categorical variable and get the logp Questions	2	1105	October 26, 2018
pm.Categorical for matrix of probabilities Questions	3	582	February 22, 2019
Trouble specificying X \| a, b, c, d ~ Categorical( . ) Questions	5	504	March 2, 2019