Date:         Wed, 27 May 2009 10:07:51 -0700
Reply-To:     rkamalakar73@GMAIL.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         rkamalakar73@GMAIL.COM
Organization: http://groups.google.com
Subject:      PROC MIXED (FIXED AND RANDOM EFFECTS)
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

hI,

  I am running a mixed mode with change percent being my response
variable and age, gender and many binary variables as independent
variables. Each observation or claim pertains to one lab of a person.
A person has atleast one lab and max of 3 labs i.e. a person can have
atleast one claim or max of 3 claims. I ran the model with fixed
effect being my treatment and binary variables as random effects in
the mixed mode, but i am getting wierd results for eg. the output says
1 subject and 19000 obs per subject. and

it also says denominater degree of freedom as "13E3" . what exactly is
happeninghere?

My code:

proc mixed data=final_lab_STATIN (WHERE=(GDR_CD NE 'U'));
where lab='LDL';
class trt cad hyper_lip hypert gdr_cd;
model chg_per=trt age/solution;
random cad hypert hyper_lip gdr_cd;
lsmeans trt/ pdiff=control ('ACTOS') ADJUST=DUNNETT;
run;

Please help.