Proc REG

This procedure is more like a general purpose method of regression analysis. 

proc reg data=dataname1 

                  outest=dataname2 rsquare outseb outvif 

                  corr 

                  simple 

                  ; 

model var2=var2-6 var10 var11 var13-16 

                    / Collin 

                        vif 

                        tol 

                        selection=forward 

                        best=4 

                        noint 

                        sle=0.1

                        sls=0.05

                        start=5

                        stop=8

                        adjrsq

                        aic

                        mse rmse sse

                        corrb

                        covb

                        clb cli clm

                        p;

run;