有一道题，试了很多次都错，所以来请教下，

CPI是因变量  PPI是自变量 从1980-2006的27年的数据

现在说是越近的年份的数据就越准备的描述了数据的特征，to incorporate this idea,a greater weight is assigned to the squared errors of recent observations. for the most recent observation, assign a weight equal to(0.95)^0=1 to the previous year'sobservation,assign a weight of (0.95)^1=0.95,and so on,assigning a weight of (0.95)^n-1for the first observation in your dataset. Conduct a regression analysis using weighted least squares(include a weight statement in the regression produre,proc reg)
我不太明白 squared errors 是在output里的结果，怎么在把他按照权重做啊？还是不是很理解，请高人指点~~谢谢了