Dear Zhu, Yi:
Here is one way:
sort firm_name alliance_date
qui forval i = 1/`=_N' {
count if firm_name == firm_name [`i'] &
inrange(alliance_date-alliance_date[`i'] , 1, (5*365+1))
replace alliance_no= r(N) in `i'
}
It may take a while, as it has to run through all observations one by
one, but it should work.
Best
Gao
2009/4/21 朱琦 <[email protected]>
>
> Question
>
> Dear Stata List!
>
> I am working on a project that examines a firm's international strategic alliance (ISA).
>
> In my huge dataset, I have around 2000 firms forming ISA between 01/01/1989- 0/01/1999. Each firm is likely to form multiple international strategic alliances during this period.
> For example: Firm A forms the first strategic alliance on 01/02/1989; the second on 04/09/1990; the third on 06/09/1997.
>
> What I need to record is the number of new strategic alliances that the firm forms within 5 years after the focal one.
> For firm A, the number of new strategic alliances that Firm A forms within 5 years is 1 corresponding to its first strategic alliance on 01/02/1989;
> the number of new strategic alliances that Firm A forms within 5 years is 0 corresponding to its second strategic alliance on 04/09/1990 ;
> the number of new strategic alliances that Firm A forms within 5 years is 0 corresponding to its third strategic alliance on 06/09/1997;
>
> As it is a huge dataset, I can't do it by hand. How can I do it by using STATA? Thanks!
>
>
>
> *
> * 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/
*
* 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/
