Re: st: Teething troubles in using -gen, mod()-

[jean ries <ries@ires.ucl.ac.be>]

Clive,

I think you should sort your data by id after expanding. In fact, the 
observations created by -expand- are added to the end of the data set. 
Try this:

. expand 5
. sort id
. gen alt = mod(_n,5)

Hope this helps,

jean


Clive Nicholas wrote:


> All,
>
> Hiya! I'm attempting to create an -expand-ed dataset by creating a copy of
> every observation (-id-) for each category of a multinomial variable
> (-gewin-). There are five categories in total.
>
> The data was already carefully checked and -sort-ed. After some very
> careful reading of appropriate sources, I thought I'd cracked it with the
> following:
>
> . use mydata
>
> . expand 5
> (10496 observations created)
>
> . gen alt = mod(_n, 5)
>
> . sort id election alt
>
> Thus, each duplicated observation within each quintet should take a unique
> value of -alt-, from 0-4. Unfortunately, the new variable does not
> -generate- as it should for the vast majority of groups:
>
> . list id pano election name1997 gewin edyear alt in 1/30, sepby(id)
>
>     +--------------------------------------------------------+
>     | id   pano   election   name1997   gewin   edyear   alt |
>     |--------------------------------------------------------|
>  1. |  1      1          1   Aberavon       2     1976     0 |
>  2. |  1      1          1   Aberavon       2     1976     1 |
>  3. |  1      1          1   Aberavon       2     1976     1 |
>  4. |  1      1          1   Aberavon       2     1976     2 |
>  5. |  1      1          1   Aberavon       2     1976     3 |
>     |--------------------------------------------------------|
>  6. |  2      1          2   Aberavon       2     1983     0 |
>  7. |  2      1          2   Aberavon       2     1983     1 |
>  8. |  2      1          2   Aberavon       2     1983     2 |
>  9. |  2      1          2   Aberavon       2     1983     2 |
> 10. |  2      1          2   Aberavon       2     1983     4 |
>     |--------------------------------------------------------|
> 11. |  3      1          3   Aberavon       2     1984     0 |
> 12. |  3      1          3   Aberavon       2     1984     1 |
> 13. |  3      1          3   Aberavon       2     1984     3 |
> 14. |  3      1          3   Aberavon       2     1984     3 |
> 15. |  3      1          3   Aberavon       2     1984     4 |
>     |--------------------------------------------------------|
> 16. |  4      1          3   Aberavon       2     1987     0 |
> 17. |  4      1          3   Aberavon       2     1987     2 |
> 18. |  4      1          3   Aberavon       2     1987     3 |
> 19. |  4      1          3   Aberavon       2     1987     4 |
> 20. |  4      1          3   Aberavon       2     1987     4 |
>     |--------------------------------------------------------|
> 21. |  5      1          3   Aberavon       2     1986     0 |
> 22. |  5      1          3   Aberavon       2     1986     1 |
> 23. |  5      1          3   Aberavon       2     1986     2 |
> 24. |  5      1          3   Aberavon       2     1986     3 |
> 25. |  5      1          3   Aberavon       2     1986     4 |
>     |--------------------------------------------------------|
> 26. |  6      1          4   Aberavon       2     1988     0 |
> 27. |  6      1          4   Aberavon       2     1988     1 |
> 28. |  6      1          4   Aberavon       2     1988     1 |
> 29. |  6      1          4   Aberavon       2     1988     2 |
> 30. |  6      1          4   Aberavon       2     1988     3 |
>     +--------------------------------------------------------+
>
> If I'm missing something obvious here, it's very esoteric.
>  
*
*   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/