Re: st: cross grouping a family w siblings and stepsiblings

[Tinne Steffensen <tst@sfi.dk>]

Thank you for the link. The problem is I don't have any family id - as have all the examples in the link. The reason is, that I'm both interested in stepsiblings on the mother's and father's side, which makes it impossible to make distinct families. (Some children will be part of two families).

Instead I will try to go back to the datasets prior to merging;
It looks like this, where group_m is (by m_id, sort: gen group_m = _n).

m_Id    child_id    gender    group_m                                  birth        handicap

1               24            2                    1                  01jan07         1
1               12            2                    2                  02feb08         0
2               47            1                    1                  10jun02         0
3               10            2                    1                  02apr00         0
4                8            2                    1                  18dec09         1
4               18            1                    2                  11nov02         1
4               56            1                    3                  18may99         0
5               82            1                    1                  23jun08         0
6               12            2                    1                  30jul01         0

I want it to come out like below.

  m_Id    child_id1  child_id2    child_id3   birth1      birth2        birth3  

1             24        12            .        01jan07    02feb08         .
2             47         .            .        10jun02          .         .
3             10         .            .        02apr00          .         .
4              8        18           56        18dec09    11nov02     18may99
5             82         .            .        23jun08          .         .
6             12         .            .        30jul01          .         .

I tried reshaping: 
reshape wide child_id, i(m_id) j(gr_m)

But it comes up with this error-message: 
(note: j = 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20)
Birth not constant within m_id
-
In the real dataset there are up to 20 children to each mother.

Thank you for your time. Tinne



*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/