How to get duplicates from multiple columns

[angeliquec]

Hi, I would like to detect duplicate rows in my dataset below 

data HAVE;
infile datalines dlm="|";
input COL1 :$5. COL2 :$5. COL3 :$5.;
datalines;
CAT|DOG|ANT
ANT|CAT|DOG
HORSE|CAT|HORSE
;
run;

*It's in Row 1 and 2.

/*Output Duplicates*/
data WANT;
infile datalines dlm="|";
input COL1 :$5. COL2 :$5. COL3 :$5.;
datalines;
CAT|DOG|ANT
ANT|CAT|DOG
;
run;

I'm not sure how to do it, I was thinking of transposing, sorting, then concatenation, but it seems not efficient. Perhaps there is a better way to do it. Thank you!

[Reeza · Accepted solution]

You can sort inline so to speak. 

If you can have different cases (ie Ant vs ANT) they will not show up as duplicates.

Something like this, untested code:

data temp;
set have;
call sortc(col1, col2, col3);
run;

proc sort data=temp out=want nouniquekey;
by col1 col2 col3;
run;

---

@angeliquec wrote:

Hi, I would like to detect duplicate rows in my dataset below 

data HAVE;
infile datalines dlm="|";
input COL1 :$5. COL2 :$5. COL3 :$5.;
datalines;
CAT|DOG|ANT
ANT|CAT|DOG
HORSE|CAT|HORSE
;
run;

*It's in Row 1 and 2.

/*Output Duplicates*/
data WANT;
infile datalines dlm="|";
input COL1 :$5. COL2 :$5. COL3 :$5.;
datalines;
CAT|DOG|ANT
ANT|CAT|DOG
;
run;

I'm not sure how to do it, I was thinking of transposing, sorting, then concatenation, but it seems not efficient. Perhaps there is a better way to do it. Thank you!

---