02-06-2013 06:38 PM
I have y ~ Normal (0, theta*A(rho)), where theta is a scalar and A(rho) is a first-order autoregressive structure.
I need to estimate theta and rho.
I'm trying to use NLPNRA whithin PROC ILM, but I don´t know how to work with literal matrix in a log-likelihood.
Can someone help me, please??
Thanks a lot!