How do I add observations for consecutive days in a SAS dataset?

[raajdesaii]

Hello everyone,

I have a dataset that looks like this:

ID	Date	Days
1	20JAN2011	5
1	3FEB2011	4
1	10MAR2011	1
2	8AUG2011	10
3	12DEC2011	29
4	5JAN2013	21
5	1 MAR2014	14

I want to add new observations by ID based on the number of days. How do I get this output?

ID	Date	Days
1	20JAN2011	5
1	21JAN2011	5
1	22JAN2011	5
1	23JAN2011	5
1	24JAN2011	5
1	25JAN2011	5
1	3FEB2011	4
1	4FEB2011	4
1	5FEB2011	4
1	6FEB2011	4

and so on.....

[PaigeMiller]

Are your dates actual numeric SAS dates? Or are they character strings?

Why does you output have 6 records from the first observation and 4 records from the second observation?

[raajdesaii]

They are actual numeric SAS dates. 

The first patient has 6 records where as the second patient only has 4 records and so on. Date is the variable which indicates when they started the drug and days indicates for how many dates. So patient 1 started the first treatment  on a particular date for x number of days. Same patient started another treatment on another date for x number of days. I just want to add all the days to the date so I get the dataset as I have posted.

[PaigeMiller]

If the first patient had 5 records in your output, and the second patient had 4 records in your output, then your explanation makes sense because days for the first patient is 5 and days for the second patient has 4.

But that's not what your output shows. Please explain why the first patient has 6 and the second patient has 4.

[raajdesaii]

My bad, that is just an error.

ID	Date	Days
1	20JAN2011	5
1	21JAN2011	5
1	22JAN2011	5
1	23JAN2011	5
1	24JAN2011	5
1	3FEB2011	4
1	4FEB2011	4
1	5FEB2011	4
1	6FEB2011	4

[PGStats]

data want;
set have;
do i = 1 to days;
    output;
    date = intnx("day", date, 1);
    end;
drop i;
run;

[raajdesaii]

I tried using this but it did not work, it would add the same date to all the id's. I want it to look like this:

ID Date Days
1 20JAN2011 5
1 21JAN2011 5
1 22JAN2011 5
1 23JAN2011 5
1 24JAN2011 5
1 3FEB2011 4
1 4FEB2011 4
1 5FEB2011 4
1 6FEB2011 4

[PGStats · Accepted solution]

Works for me:

data have;
input ID  Date :date9.   Days;
format date yymmdd10.;
datalines;
1   20JAN2011   5
1   3FEB2011    4
1   10MAR2011   1
2   8AUG2011    10
3   12DEC2011   29
4   5JAN2013    21
5   1MAR2014   14
;

data want;
set have;
do i = 1 to days;
    output;
    date = intnx("day", date, 1);
    end;
drop i;
run;

proc print noobs data=want; run;

....

[raajdesaii]

That worked! I had made a small error. Thanks!