pkcollapse suddenly not working as expected - problem with stat()?

[Meghan Bezerra]

Hello,

I use pkcollapse quite often to generate area under the curve and calculate incremental area under the curve. I have used it in recent days with the same code setup as always and it was functioning, however I've just gone to run the same code again and have gotten an error. Specifically, I received an error that a particular variable was already defined, and this seems to be stemming from stat() not functioning as expected; that is to say, when I run the code, more variables are being added to my dataset than what I am specifying.

I have some of my example data below, and the code I'm using is as follows:

pkcollapse time glucose, id(id) keep(fasting_glucose visit) stat(auc) trapezoid

Instead of just giving me auc as I've indicated with stat(auc), it is giving me aucline, aucexp, etc. (as if I did not specify stat() at all, so am getting all variables by default).

* Example generated by -dataex-. For more info, type help dataex
clear
input int id float(visit time glucose fasting_glucose)
3 0  0  75.5     74.2
3 0 10  76.8     74.2
3 0 15  79.6     74.2
3 0 20  88.1     74.2
3 0 30   104     74.2
4 0  0 113.5      114
4 0 10 110.5      114
4 0 15   119      114
4 0 20 135.5      114
4 0 30   174      114
5 0  0  91.6   90.625
5 0 10  92.5   90.625
5 0 15   108   90.625
5 0 20 128.9   90.625
5 0 30 159.5   90.625
6 0  0  83.1 83.52499
6 0 10 84.75 83.52499
6 0 15  87.5 83.52499
6 0 20    98 83.52499
6 0 30   129 83.52499
8 0  0 100.5   99.675
8 0 10  97.6   99.675
8 0 15 103.4   99.675
8 0 20 105.6   99.675
8 0 30   143   99.675
9 0  0  88.7     89.6
9 0 10  90.6     89.6
9 0 15    92     89.6
9 0 20  98.4     89.6
9 0 30   121     89.6
end

Any help anyone can provide would be much appreciated! Apologies in advance if this is a very silly question and I'm missing something obvious.


-Meghan

[Clyde Schechter]

I cannot replicate your difficulty on my setup (Windows, Stata 19.5 MP4):

. * Example generated by -dataex-. For more info, type help dataex
. clear

. input int id float(visit time glucose fasting_glucose)

           id      visit       time    glucose  fasting~e
  1. 3 0  0  75.5     74.2
  2. 3 0 10  76.8     74.2
  3. 3 0 15  79.6     74.2
  4. 3 0 20  88.1     74.2
  5. 3 0 30   104     74.2
  6. 4 0  0 113.5      114
  7. 4 0 10 110.5      114
  8. 4 0 15   119      114
  9. 4 0 20 135.5      114
 10. 4 0 30   174      114
 11. 5 0  0  91.6   90.625
 12. 5 0 10  92.5   90.625
 13. 5 0 15   108   90.625
 14. 5 0 20 128.9   90.625
 15. 5 0 30 159.5   90.625
 16. 6 0  0  83.1 83.52499
 17. 6 0 10 84.75 83.52499
 18. 6 0 15  87.5 83.52499
 19. 6 0 20    98 83.52499
 20. 6 0 30   129 83.52499
 21. 8 0  0 100.5   99.675
 22. 8 0 10  97.6   99.675
 23. 8 0 15 103.4   99.675
 24. 8 0 20 105.6   99.675
 25. 8 0 30   143   99.675
 26. 9 0  0  88.7     89.6
 27. 9 0 10  90.6     89.6
 28. 9 0 15    92     89.6
 29. 9 0 20  98.4     89.6
 30. 9 0 30   121     89.6
 31. end

. 
. pkcollapse time glucose, id(id) keep(fasting_glucose visit) stat(auc) trapezoid
......
. 
. list, noobs clean

    id   visit   fastin~e   auc_gl~e  
     3       0       74.2    2532.25  
     4       0        114     3877.5  
     5       0     90.625       3456  
     6       0   83.52499   2868.625  
     8       0     99.675     3258.5  
     9       0       89.6       2926

I would suggest doing the usual things in a circumstance were you are getting a result that others cannot replicate:

1. -update all, force- to make sure that you have the most recent version of everything in Stata, and that the various resources used are all mutually compatible. Then try again.

2. Reboot your computer and try again.

3. Uninstall and re-install Stata, then -update all, force-, and try again.

If none of that works, contact Stata technical support.