如何基于sas中的列合并表[重复]

Question

这个问题在这里已有答案：

• SAS merging multiple tables 2回答

我在SAS中有多个表，我想将它们合并为一个列

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

把它们合并为一个我可以做的事情

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;

但在我的实际情况中，我的表格在15-20附近，每张表中有超过5K的记录，因此这种方法变得非常慢，需要5-7分钟才能处理。

我正在寻找任何更好，更快的方法将表合并为一个列作为一个表。

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：