| Have: | |||
| ID | v1 | v2 | v3 |
| 1 | -3 | a | a |
| 1 | -2 | b | b |
| 1 | -1 | c | b |
| 1 | 1 | d | b |
| 2 | -20 | a | a |
| 2 | -6 | b | b |
| 2 | -3 | c | b |
| 2 | -2 | d | b |
| 3 | 1 | a | a |
| ID | v1 | v2 | v5 |
| 1 | -3 | a | a |
| 1 | -2 | b | a |
| 1 | -1 | c | a |
| 1 | 1 | d | a |
| 2 | -20 | a | a |
| 2 | -6 | b | a |
| 2 | -3 | c | a |
| 2 | -2 | d | a |
| 3 | 1 | a | a |
I am wondering how to create a new variable v5, if any of the values of v3 per subject are 'a' then v5 should be populated with the same value 'a' (v5 contains only a in output).?
proc sort
data=have (keep=id v3 where=(v3 = 'a'))
out=help (rename=(v3=v5))
nodupkey
;
by id;
run;
data want;
merge
have (drop=v3)
help
;
by id;
run;
Just one of the many possible solutions
data want;
do until(first.id);
set have;
by id;
v5=v2;
end;
do until(first.id);
set have;
by id;
output;
end;
run;
data have; infile cards expandtabs truncover; input ID v1 v2 $ v3 $; cards; 1 -3 a a 1 -2 b b 1 -1 c b 1 1 d b 2 -20 a a 2 -6 b b 2 -3 c b 2 -2 d b 3 1 a a ; run; proc sql; select *,case when(sum(v3='a') ne 0) then 'a' else ' ' end as v5 from have group by id; quit;
April 27 – 30 | Gaylord Texan | Grapevine, Texas
Walk in ready to learn. Walk out ready to deliver. This is the data and AI conference you can't afford to miss.
Register now and save with the early bird rate—just $795!
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.