1# xujiannk 求大侠指教啊

proc reg data=a;
model y=x z p;
output out=b r=c;
run;
data b;
set b;
c=c*c;
run;
proc univariate data=a.b;
var y;
output out=a.c sum=sum;
run;

when you get the proc means you will find the sum value of  the sum square of deviation.
also, you can directly use followings:

proc reg data=a;
model y=x z p;
output out=b r=c;
run;
proc univariate data=a.b;
var c;
output out=a.c uss=d;
run;

the 'd' is what you are looking for.

对吗