Build your own hierarchical model
In CARlasso: Conditional Autoregressive LASSO

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

We also offer two low-level one-step sampling functions rCARAlasso_ and rCARlasso_ for adaptive and non-adaptive version of the CARlasso, see reference for details.

The typical usage would be one writes their own sampler to sample the Normal latent variable and feed it to the one-step sampling. Note that these two functions will update the current states in memory directly.