Since you know it , why not try by yourself?

The sql solution will be Cartesian product joins. There is no equivalent merge solution. But a point direct access data set point will do it as showed below.


data t1 t2;
v1='a'; v2 = 1 ;output t1;
v1='a'; v2= 2 ;output t1;
v1='a'; v2= 3 ;output t2;
v1='a'; v2= 4 ;output t2;
run;

proc sql;
  create table t3 as
  select t1.v1 ,  t1.v2,  t2.v2 as v3
  from t1 , t2
  order by 2,3
  ;
quit;

data t4;
  set t1;
    do i=1 to nobs;
           set t2(rename=(v2=v3)) point=i nobs=nobs;
           output;
        end;
run;

呵呵，看来我对direct access data 的机制 认识还是太粗。

我会sql,但是sas的data步里面不会，而sql又没办法全满足剩下的要求。。。