Simulations in SAS

Interaction simulation

%let ErrorVariance=1000;
%let BlockVariance=3000;
%let block=18;
%let nsim=1;
data factor;
length Treatment $30;
call streaminit(123);
    do isim = 1 to ≁
        do block=1 to █
        rndBlock=rand("Normal",0,sqrt(&BlockVariance)); /* create block specific deviates */
            do Protein=0 to 1 by 1;
                do Moisture=0 to 1 by 1;
                     adg0042=550*Protein + 325*Moisture + -400*Protein*Moisture + rndBlock + rand("Normal",0,sqrt(&ErrorVariance));
                    if Moisture=0 & Protein=0 then Treatment="Low Moisture / Low Protein";
                    if Moisture=0 & Protein=1 then Treatment="Low Moisture / High Protein";
                    if Moisture=1 & Protein=0 then Treatment="High Moisture / Low Protein";
                    if Moisture=1 & Protein=1 then Treatment="High Moisture / High Protein";
                    output;
            end;
        end;
    end;
end;
ods graphics / imagefmt=svg;
proc sgplot data=factor;
vbox adg0042 / category=protein group=moisture;
title "Boxplot showing levels of Moisture & Protein ";
run; 
proc sgplot data=factor;
vbox adg0042 / group=Treatment;
title "Boxplot showing four Treatments ";
run;
proc mixed data=factor;
class protein moisture block;
model adg0042=protein|moisture / s cl ddfm=kenwardroger;
store mixmodel;
run;
proc plm restore=mixmodel;
effectplot interaction(x=moisture sliceby=protein) / connect clm ilink;
effectplot interaction (x=protein plotby=moisture) / connect ilink clm;
slice protein*moisture / sliceby=moisture pdiff=all means lines cl ilink adjust=tukey plots=diffogram;
lsmeans protein*moisture / lines ilink;
run;