Duplicates being tagged without being duplicates

[Juan Gonzalex]

Hi, I have encountered a problem that I do not understand. Somehow, Stata is tagging duplicates that, at first sight, are not duplicates. I have used the following command:

duplicates tag month year cusip, gen(flag)

As can be seen in the following sample, it tags these observations while they have different dates (same cusip):

* Example generated by -dataex-. For more info, type help dataex
clear
input long reprisk_id float(month year) str6 cusip byte flag
101829  9 2015 "466112" 1
101829  1 2019 "466112" 1
101829 12 2020 "466112" 1
101829  8 2019 "466112" 1
101829  9 2019 "466112" 1
101829  2 2020 "466112" 1
101829 12 2019 "466112" 1
101829  4 2018 "466112" 1
101829  3 2018 "466112" 1
101829 11 2019 "466112" 1
end

Why could this be the case?

Thanks.

[Nick Cox]

You are showing us results from some larger dataset, but the example data themselves do not show the problem.

* Example generated by -dataex-. For more info, type help dataex
clear
input long reprisk_id float(month year) str6 cusip byte flag
101829  9 2015 "466112" 1
101829  1 2019 "466112" 1
101829 12 2020 "466112" 1
101829  8 2019 "466112" 1
101829  9 2019 "466112" 1
101829  2 2020 "466112" 1
101829 12 2019 "466112" 1
101829  4 2018 "466112" 1
101829  3 2018 "466112" 1
101829 11 2019 "466112" 1
end

duplicates tag month year cusip, gen(tag)
     +-----------------------------------------------+
     | repris~d   month   year    cusip   flag   tag |
     |-----------------------------------------------|
  1. |   101829       9   2015   466112      1     0 |
  2. |   101829       1   2019   466112      1     0 |
  3. |   101829      12   2020   466112      1     0 |
  4. |   101829       8   2019   466112      1     0 |
  5. |   101829       9   2019   466112      1     0 |
     |-----------------------------------------------|
  6. |   101829       2   2020   466112      1     0 |
  7. |   101829      12   2019   466112      1     0 |
  8. |   101829       4   2018   466112      1     0 |
  9. |   101829       3   2018   466112      1     0 |
 10. |   101829      11   2019   466112      1     0 |
     +-----------------------------------------------+

.
My guess is that you'e not seeing the duplicates because the data are jumbled. So check out for example

list if cusip == 466112 & year == 2015 & month == 9

[Juan Gonzalex]

You are totally right, thank you so much!