I have a dataset with multiple timepoints and every visit has timepoint collected 3 times and if there is any timepoint that is collected more than 3 times it should come in output. I had written below code but it is not working. I had transposed all timepoints and got them into one variable called timepoint.
data abc_1;
set abc;
by SubjectId visit timepoint;
Seq+1;
if first.timepoint then Seq=1;
if seq>3 then flag=1;
run;
example dataset as below
data ndsn;
infile datalines;
input subject visit $4. TP1 TP2 TP3;
datalines;
101 Day1 1 1 1
101 Day1 1 2 2
101 Day1 3 3 3
101 Day1 4 4 4
101 Day1 7 7 7
101 Day1 7 8 8
101 Day1 9 9 9
101 Day1 9 10 10
101 Day1 11 11 11
101 Day2 1 1 1
101 Day2 2 2 2
101 Day2 2 3 3
101 Day2 4 4 4
101 Day2 5 5 5
101 Day2 5 6 6
101 Day2 7 7 7
101 Day2 10 10 10
101 Day2 11 10 11
;
run;
desired output as below after transposing the 3 variables to one variable called timepoint i got that variable
data ndsn;
infile datalines;
input subject visit $4. timepoint;
datalines;
101 Day1 1
101 Day1 7
101 Day1 9
101 Day2 2
101 Day2 5
101 Day2 10
;
run;
Any help please
Hi,
you could use SQL to achieve this:
PROC SQL;
CREATE TABLE ndsn_t AS
SELECT subject,visit,TP,count(TP) AS COUNT
FROM
(SELECT subject,visit,TP1 AS TP FROM ndsn
UNION CORRESPONDING ALL
SELECT subject,visit,TP2 AS TP FROM ndsn
UNION CORRESPONDING ALL
SELECT subject,visit,TP3 AS TP FROM ndsn)
GROUP BY subject,visit,TP
HAVING COUNT>3
ORDER BY subject,visit,TP
;
QUIT;
- Cheers -
Hi,
you could use SQL to achieve this:
PROC SQL;
CREATE TABLE ndsn_t AS
SELECT subject,visit,TP,count(TP) AS COUNT
FROM
(SELECT subject,visit,TP1 AS TP FROM ndsn
UNION CORRESPONDING ALL
SELECT subject,visit,TP2 AS TP FROM ndsn
UNION CORRESPONDING ALL
SELECT subject,visit,TP3 AS TP FROM ndsn)
GROUP BY subject,visit,TP
HAVING COUNT>3
ORDER BY subject,visit,TP
;
QUIT;
- Cheers -
Try this
data ndsn;
infile datalines;
input subject visit $4. TP1 TP2 TP3;
datalines;
101 Day1 1 1 1
101 Day1 1 2 2
101 Day1 3 3 3
101 Day1 4 4 4
101 Day1 7 7 7
101 Day1 7 8 8
101 Day1 9 9 9
101 Day1 9 10 10
101 Day1 11 11 11
101 Day2 1 1 1
101 Day2 2 2 2
101 Day2 2 3 3
101 Day2 4 4 4
101 Day2 5 5 5
101 Day2 5 6 6
101 Day2 7 7 7
101 Day2 10 10 10
101 Day2 11 10 11
;
run;
data temp(drop = tp:);
set ndsn;
array t tp:;
do over t;
timepoint = t;
output;
end;
run;
proc summary data = temp nway;
class subject visit timepoint;
var timepoint;
output out = want(where = (_freq_ > 3)) n =;
run;
Although all the suggestions were useful. @Oligolas i had used your suggestion in my program and it worked for me. thanks a lot for your time everyone
SAS Innovate 2025 is scheduled for May 6-9 in Orlando, FL. Sign up to be first to learn about the agenda and registration!
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.