Re: st: row wise summation in panel data for particular IDs

[pb2442@columbia.edu]

Forgot to add what I have already tried.
I used:

***
gen c=0
replace c=1 if id==1 | id==2
collapse (sum) var1, by (c year)
egen newid=group(id)
***

But, last line does not run, as old 'id' is already cleard. And, other  
variables, var2, var3...etc also do not show up any more.

I think there must be some other easier and more flexible way.


thx,
prabhat


Quoting pb2442@columbia.edu:

> Dear all,
>
> I am required to take into account the changes in boundary of cities
> over a period so that all the information can be attributed to the
> parent cities at the start of the time period. I have already tried but
> could not find a similar case in previous threads.
>
> This is a panel data , say with id, year and vars. I need to merge
> (sum or weighted sum) two particular ids completely to form a new one.
> For instance, I have:
>
> id	year	var1   var2   var3
> 1	2001	40
> 1	2002	50
> 1	2003	60
> 2	2001	400
> 2	2002	500
> 2	2003	600
> 3	2001	4000
> 3	2002	5000
> 3	2003	6000
>
>
> desired Output will be like (summing up corresponding values of var1 in
> given year for id 1 and 2, keeping 3 intact and having the results in a
> newly created id, say 4).
>
> id	year	var1     var2   var3
> 1	2001	40
> 1	2002	50
> 1	2003	60
> 2	2001	400
> 2	2002	500
> 2	2003	600
> 3	2001	4000
> 3	2002	5000
> 3	2003	6000
> 4	2001	440
> 4	2002	550
> 4	2003	660
>
> Any help will be appreciated.
>
>
> Thanks.
>
> Prabhat
>
> *
> *   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/