st: quasi-GLS

[Wu Zhang <wuzhang50@yahoo.com>]

Hi, folks,

    I would like to do the following GLS for a cross sectional data: y=beta*x+e


1. I assume the covariance matrix of the error term, e, is omg is exchangeable with parameter rou up to order of 500;

2. Use the omg matrix to do GLS; since both beta and rou are unknow, this is not the usual GLS, I call it "quasi-GLS"

Could someone give me some help about how to implement it in STATA?

Thanks,

Wu 



________________________________

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/