The construct_Sigma_nms function I introduced in a previous commit, only
returned the Sigma_nms, but it was supposed to alter some of the values
in theta_L to generate the correct Sigma matrix for the random effects.
To fix the bug that created, I have just scrapped the
construct_Sigma_nms function, and reintroduced the original code to both
stan_glm.fit.R and stan_jm.R
The construct_Sigma_nms function I introduced in a previous commit, only returned the Sigma_nms, but it was supposed to alter some of the values in theta_L to generate the correct Sigma matrix for the random effects.
To fix the bug that created, I have just scrapped the construct_Sigma_nms function, and reintroduced the original code to both stan_glm.fit.R and stan_jm.R