I find that there are lots of duplicated observations in my data sets. Is there a code to delete them automatically?
proc sort data=yourdatafile nodupkey;
may do what you want.
Possibly more effecient in some situations:
Create table nodup as
select distinct *