=========================================================================
Date:         Fri, 19 Jul 2002 10:25:56 -0700
Reply-To:     Dale McLerran <stringplayer_2@YAHOO.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Dale McLerran <stringplayer_2@YAHOO.COM>
Subject:      Re: Generalized least squares with known covariance matrix
Comments: To: Bruce Bradbury <b.bradbury@UNSW.EDU.AU>
In-Reply-To:  <c64978d6.0207181722.3d25d60f@posting.google.com>
Content-Type: text/plain; charset=us-ascii

Bruce,

Take a look at the procedure MIXED.  It should handle the problem.
You have not given much information about the structure of the error
covariance so I don't know whether you should employ a RANDOM or
REPEATED statement to inform the procedure of your error structure.
My guess is that you will employ a REPEATED statement, but I cannot
be certain.  You can specify the covariance structure parameters
employing the PARMS statement and indicate that the procedure should
NOT iterate to determine parameters which maximize the likelihood.
If you know the parameters up to a scalar multiple, then you would
not want to iterate.

Dale


--- Bruce Bradbury <b.bradbury@UNSW.EDU.AU> wrote:
> I want to fit a single equation regression model where the errors are
> not
> iid, but instead are correlated, with a covariance that is known (up
> to a
> scalar multiple). What is the easiest way to do this in SAS?


=====
---------------------------------------
Dale McLerran
Fred Hutchinson Cancer Research Center
mailto: dmclerra@fhcrc.org
Ph:  (206) 667-2926
Fax: (206) 667-5977
---------------------------------------

__________________________________________________
Do You Yahoo!?
Yahoo! Autos - Get free new car price quotes