rho_samp: sampling for [rho|-] in hierchical model
In jrlewi/brlm: Bayesian Restricted Likelihood Methods

sampling for [rho|-] in hierchical model

fn.compute.SigmaRhoInv(rho, K)

fn.compute.logDetSigmaRho(rho, K)

fn.proposal.rho(rhoCur, rho_step)

fn.loglike.rho(rho, Z, alpha_rho, beta_rho)

fn.sample.rho(rhoCur, mu_rho, psi_rho, Z, rho_step)

`rho`	desc
`K`	desc
`rhoCur`	desc
`rho_step`	desc
`Z`	desc
`alpha_rho`	desc
`beta_rho`	desc
`mu_rho`	desc
`psi_rho`	desc
`Zdesc`	desc

J_nGroups must be defined, I_nGroups must be defined function to effeciently caclulate Sigma_rhoInv using its special form and a well known matrix inverse result (from the fix-rank kriging chapter 8 of the spatial handbook book); note; the K variable added for the sampling of the z_i's later. fn.compute.logDetSigmaRho used to effeciently caclulate log(det(Sigma_rho)) using its special form and Sylvester's determinant theorem: det(I_n+AB)=det(I_m+BA)

jrlewi/brlm documentation built on March 17, 2021, 1:10 a.m.