Dear all, I have T Y variables, Y1,...,YT, which are T repeated observations on a variable over the T waves of a survey. I'd like to estimate a correlation matrix cor(Y1,...,YT) assuming specific structures, as for example exchangeable, stationary, autoregressive and nonstationary. Is there any command/package in R that I could use? Thanks in advance. Marcel _________________________________________________________________ Express yourself with cool emoticons - download MSN Messenger today!