Stata is really excellent at these problems.
-egen- haters would want to do this directly.
. bysort mothid bo : gen multi = _N
Alternatively,
. egen multi = sum(1), by(mothid bo)
or
. egen multi = count(1), by(mothid bo)
or
. egen multi = count(offid), by(mothid bo)
The last is a little more subtle. -count(arg)-
counts how many times arg is not missing.
If in this case -offid- were ever missing
you would want protection against that.
Nick
[email protected]
roger webb
> I have a large vital statistics register of all births in Denmark and
> have the following variables:
>
> Offspring ID ('offid')
> Birth order ('bo')
> Maternal ID ('mothid')
>
> If the birth order number is duplicated within a sibling set,
> multiple
> births are indicated. Eg. for 3 mothers (101,102,103):
>
> mothid bo offid
> 101 1 1
> 101 2 2
> 102 1 3
> 102 1 4
> 103 1 5
> 103 1 6
> 103 1 7
>
> Mother 101 has two singleton offspring (1,2), mother 102 has twins
> (offspring 3,4) and mother 103 has triplets (offspring 5-7).
>
> I'd like to add a multiplicity indicator ('multi') to the
> data set, i.e.:
>
> mothid bo offid multi
> 101 1 1 1
> 101 2 2 1
> 102 1 3 2
> 102 1 4 2
> 103 1 5 3
> 103 1 6 3
> 103 1 7 3
>
> I've looked at the 'egen' command but can't see anything suitable.
> Does anyone have any ideas as to how I can write a Stata
> command to generate this variable?
*
* 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/
