SQL MERGE USING MACRO FOR MULTIPLE VARIABLES

[newbie_ari]

Hi,

I've 2 datasets which has 6 columns in 1 dataset & another has 8 columns. But first 5 columns are same in both. Say first dataset name is A and another B. I want to merge the 2 dataset such that I get 9 total columns. Generally we merge using SQL in the following way:

proc sql;
create table example as
select a.*,b.var1,b.var2,b.var3
from A as a left join B as b
on a.key1 = b.key1 and a.key2 = b.key2 and a.key3 = b.key3 and a.key4 = b.key4 and a.key5 = b.key5;
quit;

And if I need to merge the table again and again, I've to repeatedly write the 5 keys again and again.

My question is, is there any way I can pass the key in macro and call the macro while merging? 

Thanks!

[Reeza]

Try a Natural JOIN instead.

[newbie_ari]

@Reeza: Thanks for your reply. But could you please elaborate on what you mean by natural join.

[Reeza · Accepted solution]

Google is your friend. 

http://www.datasavantconsulting.com/roland/natjoin.html

[newbie_ari]

@Reeza: Thanks a lot! Your answer was so uncommon I didn't bother to google it. I thought you wrote something wrong. I'm working on SQL for quite sometime & had the impression that I know SQL pretty well. But life has it's own way to show you your limitations & keeps you humble. Thanks again!