I am having trouble getting any model to sample using an aesara scan. Here is a minimum example for reproduction, along with the error I am seeing. Does anyone have advice on what I am doing wrong? from aesara import scan import pymc as pm with pm.Model() as model: μ = pm.Normal("μ", 0, 2.5) …

Are you running an old version of pymc/aesara by any chance? That sounds like a bug that was solved sometime ago. I have tested locally and your snippet works for me with pymc 4.1.3

Problem using simple aesara scan in PyMC model

Questions v5

AustinRochford August 24, 2022, 12:23pm 3

Seems likely to be the reason, let me update.

Topic		Replies	Views
Cannot convert to variable (Aesara scan) Questions	1	1439	April 25, 2021
Simple markov chain with aesara scan v5 scan_ops	12	1095	April 6, 2023
Aesara/tt.scan equivalent in pymc v5? Questions	1	356	December 22, 2022
Using aesara.scan to compute the linear predictor involving a B-spline matrix v5	8	757	October 19, 2022
Aesara.scan() creating weird intermediate variables that crash the model v5 bug	4	515	September 26, 2022

Problem using simple aesara scan in PyMC model

Related topics