re: st: out of sample prediction after xtreg

[Christopher Baum <kit.baum@bc.edu>]

<>
Will wrote

> Can you just predict xb and predict u and then carry the u values
> forward from the 1990s to the 2000s?  The intercept should stay the
> same, right?  Then generate xbu =xb+u
> 
> As for how to carry the intercepts forward maybe try:
> sort u
> bysort groupid (u): replace u=u[1] if u[_n]==.
> 
> I think the logic is sound but the syntax may not be.  For me, syntax
> writing is trial and error, mostly error.  Maybe one of the resident
> experts can chime in with better advice.



An alternate solution is illustrated in the code at  

http://www.hsph.harvard.edu/cgi-bin/lwgate/STATALIST/archives/statalist.1203/Date/article-713.html

Kit

Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                             An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html
  An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html


*
*   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/