**Date:** Wed, 29 Nov 2006 17:32:23 -0600
**Reply-To:** Robin High <robinh@UNLSERVE.UNL.EDU>
**Sender:** "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
**From:** Robin High <robinh@UNLSERVE.UNL.EDU>
**Subject:** Re: how to plot this????
**In-Reply-To:** <200611291907.kATHf6cU012381@mailgw.cc.uga.edu>
**Content-Type:** TEXT/PLAIN; charset=US-ASCII
> How to plot?
>
> I have a design with three predictors, one between-subject, dichotomous,
> that I treat as a class variable, and two continuous variables that are
> measured within participants. The DV is a continuous variable. I use the
> following proc mixed:
>
> proc mixed ic;
> class subject FACTOR;
> model DV= FACTOR|continuous1|continuous1;
> repeated /type=cs sub=subject;
> run;quit;
>
> I have a 3-way interaction, and I would like to plot the interaction
> between continuous1 and continous2 separately for the two values of FACTOR
> (which is a between-subject factors). Any idea of how to do that?
>

Hi Eman,

Something like this, where you enter min and max of x1, x2, should work:

ods output lsmeans=lsm;
ods output diffs=dfs;
ods exclude diffs lsmeans;

proc mixed ic;
class subject FACTOR;
model DV= FACTOR|x1|x2;
repeated /type=cs sub=subject;

* RANDOM subject; * RANDOM statement also works like the REPEATED given;

lsmeans factor / diff at (x1 x2 )= (1.5 3); * at min x1, min x2;
lsmeans factor / diff at (x1 x2 )= (2.5 3); * at max x1, min x2;
lsmeans factor / diff at (x1 x2 )= (1.5 5.5); * at min x1, max x2;
lsmeans factor / diff at (x1 x2 )= (2.5 5.5); * at max x1, max x2;
run;

proc sort data=lsm;
by factor x1 x2;

DATA plt; SET lsm;
by factor x1;
if first.x1 then id+1;

* id variable helps keep track of factor and the extremes of x1;

proc print data=plt run;

GOPTIONS <enter options> ;

* choose favorite plotting options;

SYMBOL1 i=join value=dot line=1 h=1 color=blue;
SYMBOL2 i=join value=dot line=1 h=1 color=red;
SYMBOL3 i=join value=dot line=1 h=1 color=green;
SYMBOL4 i=join value=dot line=1 h=1 color=purple;

PROC GPLOT;
PLOT estimate * x2 = id / NOframe haxis = 2 to 6 by 1 hm=0;
format estimate 5.1 ;
run; quit;

Robin High