Adding Varying Number of Rows

[djbateman]

Sorry for the vague title.  I couldn't think of a way to summarize my task.

I have a dataset that has a subject ID with a date that they came on the study (in MONYY7. format) and the number of months that they were on the study.  I want to create a long, vertical table that has one entry for every month they were on study.  I cannot picture an "easy" method to do this.  I can only imagine storing macro variables with distinct subject IDs and number of months then looping through all subjects and number of months.  I have 321 subjects, and the months go up to 21, so looping just doesn't seem as efficient as I would like.  Are there any ideas?  I was wondering if PROC TRANSPOSE would have such capabilities.tend a D

Example Dataset:

SUBJECT	ORIGIN	MONTHS
001	JAN2012	1
002	MAR2012	4
003	APR2012	3
004	SEP2012	9

Example Outcome:

SUBJECT	MONTH
001	JAN2012
002	MAR2012
002	APR2012
002	MAY2012
002	JUN2012
003	APR2012
003	MAY2012
003	JUN2012
004	SEP2012
004	OCT2012
004	NOV2012
004	DEC2012
004	JAN2013
004	FEB2013
004	MAR2013
004	APR2013
004	MAY2013

[Haikuo · Accepted solution]

data have;

     input SUBJECT$  ORIGIN:monyy7.  MONTHS;

     format origin monyy7.;

     cards;

001 JAN2012 1

002 MAR2012 4

003 APR2012 3

004 SEP2012 9

;

data want;

     set have;

     format month monyy7.;

     do n=0 by 1 to months-1;

           month=intnx('month', origin,n);

           output;

     end;

     keep subject month;

run;