Thank you very much Martin.
I will try it.
Best regards,
Ferdinand
On 10/20/09 3:15 AM, "Martin Weiss" <[email protected]> wrote:
>
> <>
>
> So you would end up with 10 different z1-z10?
>
>
> *************
> clear*
> set obs 10
>
> gen byte industry=_n
>
> expand 100
>
> gen x1=rnormal()
> gen x2=rnormal()
> gen x3=rnormal()
> gen x4=rnormal()
>
> gen y=1+x1-2*x2+rnormal()
>
> forv i=1/10{
> reg y x1 x2 if industry==`i'
> gen z`i'=_b[_cons]+_b[x1]*x3+_b[x2]*x4
> }
> *************
>
>
>
> HTH
> Martin
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]] Im Auftrag von Ferdinand
> Siagian
> Gesendet: Dienstag, 20. Oktober 2009 06:30
> An: [email protected]
> Betreff: st: Regression by SIC and predicting different model
>
> Hi Statalist,
>
> I try to run a regression for 10 different industries and use the
> coefficients to predict a different model.
>
> For example:
>
> I regress the following regression for different industries:
>
> Y = a + b X1 + c X2 + error
>
> So, I use the following command
>
> By industries: regress Y X1 X2;
>
> I would like to use a-hat, b-hat, and c-hat to predict:
>
> Z = a-hat + b-hat X3 + c-hat X4.
>
> I tried "predict" but it did not work. I also tried using _b[var] but it did
> not work.
>
>
>
>
>
> *
> * 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/
*
* 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/
