Re: st: RE: A Previous Question on Selecting Sample from Panel Data

[Hong Nguyen <nguyen@scranton.edu>]

Hi Nick,

Thank you!

Your new commands work in that I could now use either 2000 or 2001 (or 
both years) for selecting a sample and spread the observations to the 
other years.

Hong

Nick Cox wrote:
> This is not very clear. What do you mean by "seems to work, but somehow
> not quite"? 
>
> gen byte OK = 1 if interestonly == 1 & (yearid == 2000 | yearid == 2001)
>
> bysort gvkey (OK) : replace OK = OK[1]
>
> is one way to do it. 
>
> gen byte OK == interestonly == 1 & (yearid == 2000 | yearid == 2001)
>
> bysort gvkey (OK) : replace OK = OK[_N] 
>
> is another way to do it. 
>
> Nick 
> n.j.cox@durham.ac.uk 
>
> Hong Nguyen
>
> The procedure (that is,
> bysort firm (derivuse) : replace derivuse = derivuse[1] ==1) seems to 
> work, but somehow not quite.
>
> I have 19 years (YEARID=1987, 1988, ..., 2005 already in the data set) 
> of financial data for 423 firms set up as panel data, each firm with a 
> unique GVKEY. For 2000 and 2001 only, I have a variable called 
> INTERESTONLY, which is either 1 or 0 for these 423 firms. What I want to
>
> do is to pick observations for a sample of firms where INTERESTONLY=1 
> (in 2000 only or in 2001 only, or in either), and spread them to all 
> other years.
>
> I hope this makes things clearer as to what I have and what I'd like to 
> be able to do.
>
> *
> *   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/