Creating new set with only one observation per subject from last set

[MBlack732]

this is from BIST0535.hw3_1

As you can see I have sorted data by subject and then dov, such that the observations for a single subject are all clumped together and the observation from the first time they visited (date of visit, dov) is first. I need to create a new data set BIST0535.hw3_2 which contains only one observation from each subject with the first time they visited. I think we are meant to use an IF THEN statement, however if there is a easier way I am open to it.

Obs subject dov dbp sbp

1	07/09/1998	108	157
1	09/23/1998	107	134
2	10/15/1998	114	149
2	01/14/1999	112	165
2	04/15/1999	105	158
2	07/22/1999	115	143
3	07/22/1998	103	148
3	10/07/1998	107	165
3	01/14/1999	102	159
4	01/23/2000	87	110
4	04/26/2000	85	128
5	11/10/1999	65	137
5	02/10/2000	66	108

[novinosrin · Accepted solution]

if i understand you,

all you need is 

if first.subject;

data want;

set have;

by subject dov;

if first.subject;

run;

[Jagadishkatam]

data want;
set have;
by subject;
if first.subject;
run;

[MBlack732]

Thank you! For some reason it was giving me an error about the BY statement, but i guess there was a need for a PROC SORT before this could occur.

[novinosrin]

Yes @MBlack732 By group processing does require ordering