Retain variables when reshaping long to wide

[JBCooney]

I am attempting to reshape the following long data set to a wide data set and retain individual identifying information.  The data set has missing values.  I can transpose the repeating VAR values but can't seem to carry the TRT and SEX values with the transposition.

Long Data Set

ID     TIME VAR TRT SEX

01     1     1150     1     M

01     2     1158     1     M

02     1     1201     1     F

02     2     1236     1     F

02     3     1240     1     F

03     2     1358     2     M

04     1     1285     2     F

04     2     1280     2     F

04     3     1290     2     F

Desired Wide Data Set

ID     VAR1 VAR1 VAR3      TRT SEX

01     1150     1158            .     1     M

02     1201     1236     1240     1     F

03            .     1358            .     2     M

04     1285     1280     1290     2     F

Scoured previous posts and could not find a similar example.  Thanks for any assistance.

[PGStats · Accepted solution]

Try

proc transpose data=long out=wide prefix=var;

by id trt sex notsorted;

id time;

var var;

run;

PG

[JBCooney]

Dear PGStats,

I forgot to specify sex $ when I ran the example in SAS UE.  Nevertheless, your suggestion scaled up nicely to my larger data set. The NOTSORTED option in the BY statement seems to be what I was missing.

Many thanks!!!

[Jagadishkatam]

Alternatively by arrays in datastep

Thanks,

Jag

[JBCooney]

Jag,

Thanks so much for your guidance.  After some study your reply has really helped me to better understand array processing in SAS.  Very powerful. 

Best,

John