Try
. bysort urno (date) : gen newvar = sum(date != date[_n-1])
Nick
[email protected]
Kevan Polkinghorne
> I have a dataset as shown with a unique id no (urno) and a date of an
> episode of interest. As can be seen there can be 'multiple
> episodes' for
> each unique id but also multiple episodes on the same date for each
> unique id. I want to create an identifier (new variable below) that
> labels each date as a separate episode for each patient (ie
> 1,2,3...) but
> not for the same date. For two episodes on the same date I would like
> the same number to be allocated (see below id 10069 & 166056):
>
> Data set sorted for id and date
>
> +---------------------+
> urno date new variable
> ---------------------
> 1. 29665 09 Mar 93 1
> 2. 29665 19 May 93 2
> 3. 29665 09 Aug 93 3
> 4. 29665 05 Mar 94 4
> 5. 29665 20 Apr 94 5
> ---------------------
> 6. 29665 27 Jul 94 6
> 7. 29665 07 Jan 95 7
> 8. 78602 09 Jun 92 1
> 9. 78602 07 Jul 92 2
> 10. 100069 11 Mar 99 1
> ---------------------
> 11. 100069 06 Apr 99 2
> 12. 100069 06 Apr 99 2
> 13. 131608 15 Oct 02 1
> 14. 145132 30 Aug 01 1
> 15. 145132 20 May 02 2
> ---------------------
> 16. 166056 16 Jun 92 1
> 17. 166056 26 Jul 94 2
> 18. 166056 27 Mar 95 3
> 19. 166056 09 Mar 98 4
> 20. 166056 09 Mar 98 4
> ---------------------
> I hope this all makes sense. Does anyone know a way of doing
> this? Tried
> explicit subscripting but can seem to do exactly what I want.
> Unable to
> find any clue for archives or stata FAQ.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
