Hi there,
I have the dataset "have" that record the state as follow, state=1 means the event happened, zeros otherwise.
The event may occur more than once over the follow-up time for a given subject.
I want to calculate the start and stop time for each event to get the dataset "want" as follow.
For example, the first event (state=0) of ID_1 start at 0 and end at 2.
Have
| Date | ID | state | count |
| 199801 | 1 | 0 | 1 |
| 199802 | 1 | 0 | 2 |
| 199803 | 1 | 1 | 3 |
| 199804 | 1 | 1 | 4 |
| 199805 | 1 | 1 | 5 |
| 199806 | 1 | 0 | 6 |
| 199801 | 2 | 0 | 1 |
| 199802 | 2 | 0 | 2 |
| 199803 | 2 | 1 | 3 |
| 199804 | 2 | 1 | 4 |
| 199805 | 2 | 0 | 5 |
| 199806 | 2 | 0 | 6 |
| 199807 | 2 | 0 | 7 |
| 199808 | 2 | 0 | 8 |
| 199809 | 2 | 1 | 9 |
| 1998010 | 2 | 1 | 10 |
| 1998011 | 2 | 0 | 11 |
| 1998012 | 2 | 0 | 12 |
I want to get the start and stop time of each event for each subject. And the "Newdate" is following the start time as follow.
Want
| NewDate | ID | start | stop | state |
| 199712 | 1 | 0 | 2 | 0 |
| 199802 | 1 | 2 | 5 | 1 |
| 199805 | 1 | 5 | 6 | 0 |
| 199712 | 2 | 0 | 2 | 0 |
| 199802 | 2 | 2 | 4 | 1 |
| 199804 | 2 | 4 | 8 | 0 |
| 199808 | 2 | 8 | 10 | 1 |
| 199810 | 2 | 10 | 12 | 0 |
data ISA.Have;
infile datalines missover;
input Date ID state count;
label ID='Subject ID';
datalines;
199801 1 0 1
199802 1 0 2
199803 1 1 3
199804 1 1 4
199805 1 1 5
199806 1 0 6
199801 2 0 1
199802 2 0 2
199803 2 1 3
199804 2 1 4
199805 2 0 5
199806 2 0 6
199807 2 0 7
199808 2 0 8
199809 2 1 9
1998010 2 1 10
1998011 2 0 11
1998012 2 0 12
;
Many thanks.
Regards,
Hui
