I’m trying to fit a hierarchical model using data from 30 features and 5 groups (code below). Since I do not want to define each of the 30 features’ distributions, I’m using the shape argument to speed up things. This will require to create a matrix of size (n_groups, n_features) to store the coeffi…

Hierarchical models with several of coefficients

AlexAndorra October 19, 2020, 9:30am 6

Are you sure n_stores and n_features are int and not float?
Also, try casting mu and sigma to floats: pm.Normal('b_offset', mu=0.0, sigma=1.0, shape=(n_stores, n_features))

Topic		Replies	Views
Index error in hierarchical model Questions	1	815	June 5, 2018
Hierarchical Modeling Index Error Questions	1	573	January 15, 2019
Prediction using sample_ppc in Hierarchical model Questions from_github	6	5004	December 14, 2017
Why can't I use a Bernoulli as a likelihood variable in a hierarchical model in PyMC3? Questions	2	1141	October 6, 2018
First attempt at a hierarchical model with custom likelihood Questions	1	709	September 18, 2018

Hierarchical models with several of coefficients

Related topics