初来咋到，请教一个多对多的数据合并问题！谢谢！

shiyiming

上面的output 数据库中记录：
d 7
e 8
中的7和8应是var3的数据，可能这里发帖子只有一个空格所致，大家不要误会！

再次谢谢willon,jerry!

shiyiming

proc sql;
create table c as
select coalesce(a.var1,b.var1) as var1,a.var2 ,b.var2 as var3
from a
full join b on a.var1=b.var1;
quit;

shiyiming

Excellent work!

shiyiming

proc sql;
create table output as
select coalesce(a.var1,b.var1)as var1, a.var2,b.var2 as var3
from a full join b
on a.var1=b.var1
order by var1,var2,var3;
quit;

shiyiming

其实这样写更简单,效率也更高:

data temp;
  merge a b;
  by var1;       
run;

shiyiming

merge 和sql full join的作用原理不一样
merge不会产生笛卡尔乘积，而sql会，所以应根据不同情况选择相应的语句

merge合并时虽不产生笛卡尔乘积，但也不是一一对应的，所以用merge需要比较谨慎

shiyiming

Result may not right when you merge many to many. Be careful !