I’m showing my ignorance here but, try as I might, I haven’t found the answer in the literature or in the forum. How do I specify a BYM2-style model ( Riebler et al., 2016 ) to account for spatial autocorrelation when the outcome is a continuous variable? I love how easy it is to estimate spatial con…

I think that maybe car(... type = "icar") might accomplish what you need? The key issue is not that your response is continuous, but rather that with the gaussian family the residual is confounded with a random effect of observation (if you had multiple observations per cell, then there wouldn’t be…

BYM2-style model for continuous data

Modeling

mwalks February 22, 2022, 6:57pm 6

@nkreimer did you ever end up reaching a solution on this? I referenced your post here as I am encountering similar issues.

Topic		Replies	Views
Help with picking CAR model type for hierarchical model - divergent transitions, high rhats / issues with sigma and sdcar Modeling rstan , techniques , fitting-issues , performance , brms	2	1130	February 23, 2022
Bym2 General rstan , fitting-issues , cmdstanr , hierarchical-model , brms	26	1312	February 14, 2023
Spatial conditional autoregressive (CAR) term in brms with missing observations Modeling specification	2	1288	March 22, 2021
Spatial autocorrelation brms	2	2568	July 2, 2020
Incorporating Spatial Weights Matrix brms	10	2223	April 24, 2020

BYM2-style model for continuous data

Related topics