Date:         Mon, 4 Apr 2011 20:48:18 -0400
Reply-To:     R B <ryan.andrew.black@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         R B <ryan.andrew.black@GMAIL.COM>
Subject:      Re: Nested models using NLMIXED
Comments: To: FRP3@pitt.edu
In-Reply-To:  <f7f0bcf3d6160310b38c74081e02b917.squirrel@webmail.pitt.edu>
Content-Type: text/plain; charset=ISO-8859-1

Francis,

Dale has shown how to fit a nested random effects model employing the
NLMIXED procedure, if that's what you're asking. Just search the SAS-L
archives. I'll point out, though, that it is generally much easier to
fit a nested random effects model using the GLIMMIX procedure.

I'm not sure exactly what type of joint model you're considering (I'm
actually not sure what you mean by joint model). It is possible to
model multiple dependent variables simulteanously using the GLIMMIX
procedure, even if the distributions of the dependent variables are
not the same. There are certain models that cannot be fitted using
GLIMMIX, such as mixture models like the zero-inflated poisson model.
The NLMIXED procedure is capable of fitting these models. Again, if
you search the archives, you'll find plenty of examples by Dale.

Ryan

On Mon, Apr 4, 2011 at 3:56 PM, SUBSCRIBE SAS-L Ameribrit <FRP3@pitt.edu> wrote:
> HI Guys: a friend of mine is trying to fit a nested joint model in MATLAB
> with no joy. Is there a way to trick NLMIXED into fitting a nested model?
> I think I read somewhere that it can only do a single level of nesting but
> I wasnt sure.
>
> Thanks
> Francis
>