<>
An alternative solution, which avoids having to run a regression
first, is to use -mark- and -markout-:
mark touse
markout touse Y x-x8
foreach v of varlist Y x1-x8 {
by id: egen mean_`v'=mean(`v') if touse
}
HTH,
Glenn
Roy Wada <[email protected]> wrote:
> I want to demean variables Y, x1-x8 and then run an OLS regression of
> these variables for a panel data set for respondents. I think I need
> to be careful with the missing values in these variables. I'm not sure
> to exclude any respondent who has missing values for any of the
> variable in Y, x1-x8 when calculating the group mean for each
> respondent.
>
> Mandy
You can take advantage of -reg- to perform casewise deletion.
qui reg y x-x8
foreach v of varlist Y x1-x8 {
by id: egen mean_`v'=mean(`v') if e(sample)==1
}
Roy
*
* 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/
