Setting up a Gibbs Sampler for Multivariate Normal Distribution

Good evening,

I am trying to write a code for Gibbs Sampling, for my master thesis. Unfortunately I have litlle to no experience in programming or writing code. Hence, I ran into some troubles.

I have three variables that need to be considered i call them x,y,z for simplicity.

In order to sample from the conditional distribution I need the mean vector and Covarince Matrix. However, i do not know how to write the vectors which are contaning 1x2 matrices.

So for example the mean vector would be {\mathbf {x}}={\begin{bmatrix}{\mathbf {mu}}_{1}\\{\mathbf {mu}}_{2}\end{bmatrix}}{\text{ with sizes }}{\begin{bmatrix}q\times 1\\(N-q)\times 1\end{bmatrix}}. How can i create in the first line my mu1 where I have an 1x1 "matrix" and in the second row my 2x1 matrix with mu2 and mu3?

Maybe someone could help me figuring this problem out?

Thanks in advance!

Best, Lukas

Answer 1

[Sindhu Karri]

Hii,

According to my understanding of the question ,I suppose ,you want to create a variable whose elements are matrices.This can be done by using structures.You can use struct function to create structures, mean function to calculate mean,cov function to calculate covariance matrix.