Re: st: Assigning id's to other observations (after expanding)

[Nick Cox <njcoxstata@gmail.com>]

The last two need the extra condition

... & toexpand

On Tue, Apr 19, 2011 at 6:48 PM, Nick Cox <njcoxstata@gmail.com> wrote:
> Not sure I follow this at all, but example code:
>
> gen toexpand = id == parentid
> expand 2 if toexpand
> bysort parentid toexpand (id) : gen first = id[1]
> by parentid toexpand (id) : gen last = id[_N]
> by parentid : replace first = first[1]
> by parentid : relace last = last[1]
> by parentid : replace id = first if _n == _N - 1
> by parentid : replace id = last if _n == _N
>
>
>
> On Tue, Apr 19, 2011 at 6:24 PM, Ibarra-Caton, Marilyn
> <marilyn.ibarra-caton@bea.gov> wrote:
>> I am having a hard time with coding to get the data as I need it.  Any
>> help would be greatly appreciated!!!
>>
>> My data looks like this:
>>      id          Parent id
>> 1.    Firm1   Firm3
>> 2.    Firm1   Firm3
>> 3.    Firm2   Firm3
>> 4.    Firm3   Firm3
>> In words what the data is telling me is that the parent id is the firm
>> that absorbed the firm in the id column.  So in this case firm 3
>> absorbed (merged with) firm 1 and firm 2.
>>
>> So I want the data to look like this:
>>
>> 1.    Firm1    Firm3
>> 2.    Firm1    Firm3
>> 3.    Firm2    Firm3
>>
>> 4.    Firm1    Firm3
>> 5.    Firm2    Firm3
>>
>> So basically duplicate observation 4 and assign the id of observation 1
>> & 2 to one of the duplicate observations and assign the id of
>> observation 3 to the other duplicate observation.  For the observation
>> that I am going to duplicate and assign a new id, it will always be the
>> case that the id==parent id for this observation before any manipulation
>> of the data.   I know that to duplicate observation 4 I can use the
>> expand command.
>>
>

*
*   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/