Much better… but can’t you sort things in x_r directly in a column major format so that you can do
normal_lpdf(xs[...] | to_matrix(xs[...], J, ??) * beta, sigma )
(sorry, leaving some things unclear due to lack of time to look more closely).
Much better… but can’t you sort things in x_r directly in a column major format so that you can do
normal_lpdf(xs[...] | to_matrix(xs[...], J, ??) * beta, sigma )
(sorry, leaving some things unclear due to lack of time to look more closely).