随便附个程序吧：
proc reg data=ty outest=ww;
model vol =fac stu sat/ridge=0 to 0.3 by 0.01 0.3 to 1 by 0.1;
plot/ridgeplot;
proc print data=ww;

All are in SAS proc reg.

要求保持变量个数并解决多重共线性较好的方法是岭回归吗？The ridge is an approach to trade off between unbiasness and regresion statbilities. It is not easy to judge. It is most often than not to judge it based on business intuition.