Re: st: how to create a loop for all the feasible regressions in a varlist

[Stefano Lombardi <ste85@fastwebnet.it>]

Il 23/03/2011 23:17, Nick Cox ha scritto:
> This came up only a few days ago.
>
> See http://www.stata.com/statalist/archive/2011-03/msg00770.html
>
> local varlist var1 var2 var3 var4
> tokenize "`varlist'"
>
> forval i = 1/4 {
>         local x : list varlist - `i'
>         regress ``i'' `x'
> }
>
> On Wed, Mar 23, 2011 at 9:42 PM, Stefano Lombardi<ste85@fastwebnet.it>  wrote:
>
> > Starting from a list of 4 variables, I need to:
> >     - regress var1 on var2, var3, var 4
> >     - regress var2 on var1, var3, var4
> > and so on..
> > Basically I need to create a loop where every time the dependent variable is
> > regressed on all the three others.
> >
> > I have been trying to use the "foreach" command, but I cannot find the right
> > way to solve the issue. In fact if I type:
> >
> >   foreach x in Gasp PNC PUC PPT {
> >     reg `x' Gasp PNC PUC PPT
> >     }
> >
> > the software, obviusly, every time returns an erroneous model (R^2 = 1),
> > because in each model the dependent variable is explained by itself..
> > The point is that I have no idea on how to remove the `x' variable from the
> > regressors list.
> *
> *   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/
Thank you very much!

Best regards,

Stefano Lombardi
*
*   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/