I’ve had a look around, but have not found a solution for this. I’m using a MvNormal as a prior for other parameters in a model. I’ve got this working where you manually specify the mu and cov, using pm.MvNormal. mu = np.array([-1, 0.5]) cov = np.array([[1.0, -0.5], [-0.5, 2]]) VALS = pm.MvNormal("…

Hyperpriors for a MvNormal prior

drbenvincent February 18, 2021, 5:45pm 3

Not not happy - just completely new to the LKJ so don’t have a good feel for how to change parameters to reflect different prior beliefs.

Topic		Replies	Views
Multivariate normal distribution Questions	4	2617	March 29, 2019
Using a Multivariate Normal as a prior Questions	1	1171	March 4, 2019
Priors for covariance matrix for multivariate normal Questions	3	624	August 18, 2020
Using LKJ priors Questions	3	553	December 14, 2020
Using multivariate normal with widely varying means Questions	3	517	October 1, 2019