Re: AW: AW: st: AW: mfxrcspline : error message in Stata11

[SCHOUMAKER Bruno <bruno.schoumaker@uclouvain.be>]

Thanks !

_rmcoll removes collinear variables if the option forcedrop is used 
(otherwise they are flagged o.varname).

It now works with the following syntax

***
_rmcoll `varlist'==1, forcedrop
local xvars `r(varlist)'
poisson _d `xvars'
mfxrcspline, link(log) level(80) generate(marg margl margu)
***

Bruno


Le 15/07/2010 13:41, Martin Weiss a écrit :
> <>
>
>
> " So I guess I need to find a way of dropping the collinear variables
> variables before running the models (or use Stata10)."
>
>
>
> Try
>
> *************
> help _rmcoll
> *************
>
>
> Re your attempts to -version- your way out of the problem, see NJC`s recent comment at http://www.stata.com/statalist/archive/2010-07/msg00180.html
>
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von SCHOUMAKER Bruno
> Gesendet: Donnerstag, 15. Juli 2010 13:02
> An: statalist@hsphsun2.harvard.edu
> Betreff: Re: AW: st: AW: mfxrcspline : error message in Stata11
>
> Thank you for your replies !
>
> Below is a copy of the part around the error. I noticed that dumy102 and
> bin103 had been transformed into o.dumy102 and o.bin103.
>
> In the previous version, these variables were dropped automatically from
> the regression model when there was collinearity, but it is no longer
> the case in stata11 .
>
> "As of Stata 11, variables are no longer dropped because of
> collinearity. Instead, these variables are omitted and are labeled with
> the “o.” operator in the column names of the resulting parameter
> vector.", (http://www.stata.com/support/faqs/lang/post.html).
>
> This is the cause of the problem. When I drop the collinear variables
> before running the model, it works (adding "version 10.1:" before the
> regression model and/or mfxrcspline does not solve the problem)
>
> So I guess I need to find a way of dropping the collinear variables
> variables before running the models (or use Stata10).
>
> Best,
>
> Bruno
>
>
>
>
> ***********
>       - local other : list other - notis
>       - if "`other'" != "" {
>       = if "_Iageg_5 _Iageg_6 _Iageg_7 _Iageg_8 _Iageg_9 _Iageg_10 rcsp1
> rcsp2 rcsp3 rcsp4 rcsp5 dumy82 dumy8
>   >  7 dumy94 o.dumy102 bin82 bin87 bin94 bin102 bin83 bin88 bin95
> o.bin103" != "" {
>       - foreach var of varlist `other' {
>       = foreach var of varlist _Iageg_5 _Iageg_6 _Iageg_7 _Iageg_8
> _Iageg_9 _Iageg_10 rcsp1 rcsp2 rcsp3 rcsp4
>   >   rcsp5 dumy82 dumy87 dumy94 o.dumy102 bin82 bin87 bin94 bin102 bin83
> bin88 bin95 o.bin103 {
> o. operator not allowed
>         sum `var' if `touse', meanonly
>         qui drop `var'
>         qui gen double `var' = r(mean)
>         }
>         }
> ***********
>
> Le 15/07/2010 11:36, Martin Weiss a écrit :
>    
> > <>
> >
> >
> > And yet, strangely, Bruno does not seem to have used the "factor variable notation" in any way in his call to -ssc d mfxrcspline-. I tried to recreate his problem, but all efforts failed. The "click-to-run" examples in the help file do work as expected. So I would ask Bruno to -set tr on- and report the area around the error...
> >
> >
> >
> > HTH
> > Martin
> >
> >
> > -----Ursprüngliche Nachricht-----
> > Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Maarten buis
> > Gesendet: Donnerstag, 15. Juli 2010 11:31
> > An: statalist@hsphsun2.harvard.edu
> > Betreff: Re: st: AW: mfxrcspline : error message in Stata11
> >
> > --- SCHOUMAKER Bruno wrote:
> >
> >      
> > > I used mfxrcspline with Stata10. I recently upgraded to
> > > Stata11, ran on Stata11 the do file that worked (and
> > > still works) with Stata10, and I
> > > got the following error message
> > >
> > > . mfxrcspline, link(log) level(80) generate(marg margl
> > > margu)
> > > o. operator not allowed
> > > r(101);
> > > r(101);
> > >
> > > Do you know what this message means and how to solve it ?
> > >
> > >        
> > --- On Thu, 15/7/10, Martin Weiss wrote:
> >
> >      
> > > The -program- sets -version- to 10. So the new features
> > > under 11, in particular -fvvarlist-s, are not yet
> > > implemented.
> > >
> > >        
> > Martin maybe on to something. None of my programs allow the new
> > factor variable notation. This is not because I think that it is
> > a bad idea, but just because I haven't found the time to find
> > out how to implement it. I have tried out some examples with and
> > without the new factor variable notation, and indeed everything
> > works as expected as long as you don't use the factor variable
> > notation but errors are returned as soon as the factor varaibles
> > are used.
> >
> > Thanks for the report,
> > Maarten
> >
> > --------------------------
> > Maarten L. Buis
> > Institut fuer Soziologie
> > Universitaet Tuebingen
> > Wilhelmstrasse 36
> > 72074 Tuebingen
> > Germany
> >
> > http://www.maartenbuis.nl
> > --------------------------
> >
> >
> >
> >
> >
> > *
> > *   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/
> >
> >      
>
>    


--
Bruno SCHOUMAKER

Centre de recherche en démographie et sociétés
Université catholique de Louvain
1-17 PLace Montesquieu
1348 Louvain-la-Neuve
BELGIUM

Tel. +32 10 474136
Fax. +32 10 472952

bruno.schoumaker@uclouvain.be
www.uclouvain.be/demo

*
*   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/