data test;
    input dob : yymmdd8. race $;
        format dob yymmdd10.;
    datalines;
19550101 a
19560102 b
19560102 b
19570103 c
19570103 d
19580104 e
19590105 f
19600106 g
19600106 g
;

*** Solution 1;
proc sort data=test out=aaa dupout=bbb nodupkey ;
     by dob race;
run;

*** Solution 2;
proc sql;
     create table ccc as
            select distinct dob,race
                  from test
                  group by dob,race
                  having count(*)>1;
quit;

*** Solution 3;
data ddd;
     set test;
         by dob race;
         if not first.race;
run;

Ron Cody 的Data.Cleaning.Techniques.Using.SAS有詳細講解

use DUPOUT

PROC SORT DATA=dedupe_demo NODUPKEY DUPOUT=nodup_by_key  OUT=nodup_by_key ;
  BY date_of_birth race;
RUN;