How to remove duplicate observations based on the value of the the third variable?

[zjppdozen]

Hi all,

I have a data set like this (see below). There are duplicate observations within some survey waves. Now I want to remove the duplicate observations with the highest "visit age" from the sample. What should I do?

The original data set:

ID	WAVE	Visit age
1231	1	45
1231	2	46
1231	2	47
1232	1	55
1233	1	56
1234	1	34
1234	1	35
1234	2	38

The dataset that I want:

ID	WAVE	Visit age
1231	1	45
1231	2	46
1232	1	55
1233	1	56
1234	1	34
1234	2	38

Thank you very much!

[data_null__ · Accepted solution]

---

@zjppdozen wrote:

Hi all,

I have a data set like this (see below). There are duplicate observations within some survey waves. Now I want to remove the duplicate observations with the highest "visit age" from the sample. What should I do?

The original data set:

ID	WAVE	Visit age
1231	1	45
1231	2	46
1231	2	47
1232	1	55
1233	1	56
1234	1	34
1234	1	35
1234	2	38

The dataset that I want:

ID	WAVE	Visit age
1231	1	45
1231	2	46
1232	1	55
1233	1	56
1234	1	34
1234	2	38

Thank you very much!

---

proc summary nway missing;
   class id wave;
   output out=deduped(drop=_type_) min('visit age'n)=;
   run;

[zjppdozen]

This code works! Thank you very much!