How to remove IDs with deuplicates entirely without keeping any of them in the dataset?

[Denali]

Hi,

My dataset contains duplicate, triplicate or even more subject IDs. However, I do not know which duplicate is correct so I would like to remove the IDs with more than 1 count completely.

For example, there are two subject ID#1002. I would like to remove both subject ID#1002 completely. I would also like to export all the subjects who have been removed to a dataset so I can keep track of who has duplicates in the dataset. 

Thank you all in advance!

[novinosrin · Accepted solution]

will this help?

proc sort data=have;
by id;
run;

data unique NOUNIQUEKEY;
set have;
by id;
if first.id and last.id then output unique;
else output NOUNIQUEKEY;
run;