I am trying to generate a counter variable that describes the duration of a temporal episode in panel data. I am using long format data that looks something like this:
clear
input byte id int time byte var1 int aim1
1 1 0 .
1 2 0 .
1 3 1 1
1 4 1 2
1 5 0 .
1 6 0 .
1 7 0 .
2 1 0 .
2 2 1 1
2 3 1 2
2 4 1 3
2 5 0 .
2 6 1 1
2 7 1 2
end
I want to generate a variable like aim1 that starts with a value of 1 when var1==1, and counts up one unit with each subsequent observation per ID where var1 is still equal to 1. For each observation where var1!=1, aim1 should contain missing values.
I already tried using rangestat (count) to solve the problem, however the created variable does not restart the count with each episode:
ssc install rangestat
gen var2=1 if var1==1
rangestat (count) aim2=var2, interval(time -7 0) by (id)
Here are two ways to do it: (1) from first principles, but see this paper for more and (2) using tsspell from SSC.
clear
input byte id int time byte var1 int aim1
1 1 0 .
1 2 0 .
1 3 1 1
1 4 1 2
1 5 0 .
1 6 0 .
1 7 0 .
2 1 0 .
2 2 1 1
2 3 1 2
2 4 1 3
2 5 0 .
2 6 1 1
2 7 1 2
end
bysort id (time) : gen wanted = 1 if var1 == 1 & var1[_n-1] != 1
by id: replace wanted = wanted[_n-1] + 1 if var1 == 1 & missing(wanted)
tsset id time
ssc inst tsspell
tsspell, cond(var1 == 1)
list, sepby(id _spell)
+---------------------------------------------------------+
| id time var1 aim1 wanted _seq _spell _end |
|---------------------------------------------------------|
1. | 1 1 0 . . 0 0 0 |
2. | 1 2 0 . . 0 0 0 |
|---------------------------------------------------------|
3. | 1 3 1 1 1 1 1 0 |
4. | 1 4 1 2 2 2 1 1 |
|---------------------------------------------------------|
5. | 1 5 0 . . 0 0 0 |
6. | 1 6 0 . . 0 0 0 |
7. | 1 7 0 . . 0 0 0 |
|---------------------------------------------------------|
8. | 2 1 0 . . 0 0 0 |
|---------------------------------------------------------|
9. | 2 2 1 1 1 1 1 0 |
10. | 2 3 1 2 2 2 1 0 |
11. | 2 4 1 3 3 3 1 1 |
|---------------------------------------------------------|
12. | 2 5 0 . . 0 0 0 |
|---------------------------------------------------------|
13. | 2 6 1 1 1 1 2 0 |
14. | 2 7 1 2 2 2 2 1 |
+---------------------------------------------------------+
The approach of tsspell is very close to what you ask for, except (a) its counter (by default _seq is 0 when out of spell, but replace _seq = . if _seq == 0 gets what you ask (b) its auxiliary variables (by default _spell and _end) are useful in many problems. You must install tsspell before you can use it with ssc install tsspell.