Number of repeats in each block

[Lisa NingLi]

Dear helpers
I couldn't get a good code to count the number of continuous 1s for each ID; something like
0 0 4 4 4 4 0 1 in 2013-2020 for the 1st id
2 2 0 0 0 2 2 in 2014-2020 for the 2nd id
Please help.

----------------------- copy starting from the next line -----------------------

* Example generated by -dataex-. For more info, type help dataex
clear
input int id float(yr x)
1001 2013 0
1001 2014 0
1001 2015 1
1001 2016 1
1001 2017 1
1001 2018 1
1001 2019 0
1001 2020 1
1002 2014 1
1002 2015 1
1002 2016 0
1002 2017 0
1002 2018 0
1002 2019 1
1002 2020 1
end

------------------ copy up to and including the previous line ------------------

[Daniel Schaefer]

* Example generated by -dataex-. For more info, type help dataex
clear
input int id float(yr x)
1001 2013 0
1001 2014 0
1001 2015 1
1001 2016 1
1001 2017 1
1001 2018 1
1001 2019 0
1001 2020 1
1002 2014 1
1002 2015 1
1002 2016 0
1002 2017 0
1002 2018 0
1002 2019 1
1002 2020 1
end

bysort id: egen count = total(x)

list

. list

     +-------------------------+
     |   id     yr   x   count |
     |-------------------------|
  1. | 1001   2013   0       5 |
  2. | 1001   2014   0       5 |
  3. | 1001   2015   1       5 |
  4. | 1001   2016   1       5 |
  5. | 1001   2017   1       5 |
     |-------------------------|
  6. | 1001   2018   1       5 |
  7. | 1001   2019   0       5 |
  8. | 1001   2020   1       5 |
  9. | 1002   2014   1       4 |
 10. | 1002   2015   1       4 |
     |-------------------------|
 11. | 1002   2016   0       4 |
 12. | 1002   2017   0       4 |
 13. | 1002   2018   0       4 |
 14. | 1002   2019   1       4 |
 15. | 1002   2020   1       4 |
     +-------------------------+

Edit: I'm sorry, that's not what you were asking for. I'll think about it, and I will respond with the correct answer unless someone else responds faster.

[Lisa NingLi]

Thanks, Daniel, but that's not what I wanted.
I wanted
id yr x y
1001 2013 0 0
1001 2014 0 0
1001 2015 1 4
1001 2016 1 4
1001 2017 1 4
1001 2018 1 4
1001 2019 0 0
1001 2020 1 1

[Daniel Schaefer]

Okay, I think this works:

* Example generated by -dataex-. For more info, type help dataex
clear
input int id float(yr x)
1001 2013 0
1001 2014 0
1001 2015 1
1001 2016 1
1001 2017 1
1001 2018 1
1001 2019 0
1001 2020 1
1002 2014 1
1002 2015 1
1002 2016 0
1002 2017 0
1002 2018 0
1002 2019 1
1002 2020 1
end

gen indicator = x[_n -1] == 0 & x == 1
gen sum = sum(indicator)
egen group = group(id sum) if x != 0
bysort group: gen wanted = _N if !missing(group)
replace wanted = 0 if missing(wanted)
sort id yr
drop indicator sum group

list

. list

     +--------------------------+
     |   id     yr   x   wanted |
     |--------------------------|
  1. | 1001   2013   0        0 |
  2. | 1001   2014   0        0 |
  3. | 1001   2015   1        4 |
  4. | 1001   2016   1        4 |
  5. | 1001   2017   1        4 |
     |--------------------------|
  6. | 1001   2018   1        4 |
  7. | 1001   2019   0        0 |
  8. | 1001   2020   1        1 |
  9. | 1002   2014   1        2 |
 10. | 1002   2015   1        2 |
     |--------------------------|
 11. | 1002   2016   0        0 |
 12. | 1002   2017   0        0 |
 13. | 1002   2018   0        0 |
 14. | 1002   2019   1        2 |
 15. | 1002   2020   1        2 |
     +--------------------------+

[Lisa NingLi]

Hi Daniel, great! That's exactly what I wanted! Short & nice!
Thank you so much for the help!

[Nick Cox]

Here's another way to do it using tsspell from SSC.

For the principles, see https://journals.sagepub.com/doi/pdf...867X0700700209

For examples of the practice, tsspell is an otherwise unpredictable search term for this forum.

* Example generated by -dataex-. For more info, type help dataex
clear
input int id float(yr x)
1001 2013 0
1001 2014 0
1001 2015 1
1001 2016 1
1001 2017 1
1001 2018 1
1001 2019 0
1001 2020 1
1002 2014 1
1002 2015 1
1002 2016 0
1002 2017 0
1002 2018 0
1002 2019 1
1002 2020 1
end

tsset id yr 

* ssc install tsspell 
tsspell, cond(x == 1)

egen length = max(_seq), by(id _spell) 

list, sepby(id _spell)

   +-------------------------------------------------+
     |   id     yr   x   _seq   _spell   _end   length |
     |-------------------------------------------------|
  1. | 1001   2013   0      0        0      0        0 |
  2. | 1001   2014   0      0        0      0        0 |
     |-------------------------------------------------|
  3. | 1001   2015   1      1        1      0        4 |
  4. | 1001   2016   1      2        1      0        4 |
  5. | 1001   2017   1      3        1      0        4 |
  6. | 1001   2018   1      4        1      1        4 |
     |-------------------------------------------------|
  7. | 1001   2019   0      0        0      0        0 |
     |-------------------------------------------------|
  8. | 1001   2020   1      1        2      1        1 |
     |-------------------------------------------------|
  9. | 1002   2014   1      1        1      0        2 |
 10. | 1002   2015   1      2        1      1        2 |
     |-------------------------------------------------|
 11. | 1002   2016   0      0        0      0        0 |
 12. | 1002   2017   0      0        0      0        0 |
 13. | 1002   2018   0      0        0      0        0 |
     |-------------------------------------------------|
 14. | 1002   2019   1      1        2      0        2 |
 15. | 1002   2020   1      2        2      1        2 |
     +-------------------------------------------------+