如何根据sas中的列合并表

Question

我在SAS中有多个表，其列与我想合并的列相同

table_1
obj_Id | value_txt
 1     |    T1
 2     |    T1
 3     |    T1

table_2
obj_Id | value_txt
 1     |    T2
 2     |    T2
 3     |    T2

table_3
obj_Id | value_txt
 1     |    T3
 2     |    T3
 3     |    T3

把它们合并为一体，我可以

proc sql noprint;
create table merge as
select t1.obj_id, t1.value_txt as T1, t2.value_txt as T2, t3.value_txt as T3 
from table_1 t1, table_2 t2, table_3 t3,
where t1.obj_id eq t2.obj_id
and   t1.obj_id eq t3.obj_id;
quit;

但是在我的实际场景中，我的表在15-20之间，每个表中有超过5K的记录，所以这个方法变得非常慢，需要5-7分钟的时间来处理。

我正在寻找一种更好、更快速的方法，将表合并成一个按列排列的表。

Answer 1

下面的SAS SQL和Merge代码都将创建相同的输出：

/*SAS Inner Join*/
proc sql noprint;
create table merge as select 
t1.obj_id , t1.value_txt as T1, t2.value_txt as T2, t3.value_txt as T3 
from 
table_1 t1 inner join table_2 t2 on  t1.obj_id = t2.obj_id 
inner join  table_3 t3 on t1.obj_id = t3.obj_id;
quit;
/*SAS Data Step Megrge*/
data merge_sas;
 merge table_1(rename=(value_txt=T1)) table_2(rename=(value_txt=T2)) table_3(rename=(value_txt=T3));
 by obj_id;
 run;

输出：

​

​