Pulling unique identifiers

[RobertHuey]

Hi all,

Is there a way for me to pull all values in a column that are unique?  I combined two monthly datasets and some of the primary keys are unique when combined, not all of them though.  Is there a way to pull just the unique primary keys without using PROC SQL?  I dropped some of the variables so I can't use PROC SQL according to the log.  Also, I tried using PROC SORT and the NODUPKEY.  I forgot that it keeps one set of the duplicate records but I want to get rid of both sets of the duplicate records.  Let me know of any suggestions!  Thanks!

[DBailey]

So you want to combine two datasets and then delete records that have an id with more than one occurrence?

proc sql;

delete from have

where id in (select id from have group by id having count(*)>1);

quit;

[Ksharp]

data unique;

set have;

by id;

if first.id and last.id ;

run;

and why can't use sql ?

select * from have group by id having count(*)=1  ;

Xia Keshan