Re: st: suest and factor coding

["Michael N. Mitchell" <Michael.Norman.Mitchell@gmail.com>]

Dear Tim

  Jeff Pitplado of Stata answered this question recently, see

http://www.stata.com/statalist/archive/2010-12/msg00235.html

Best regards,

Michael N. Mitchell
Data Management Using Stata      - http://www.stata.com/bookstore/dmus.html
A Visual Guide to Stata Graphics - http://www.stata.com/bookstore/vgsg.html
Stata tidbit of the week         - http://www.MichaelNormanMitchell.com



On 2010-12-17 9.25 AM, Tim Wade wrote:
> Hi Statalist,
>
> I am using Stata V 11.1 SE, Windows XP
>
> I have run across an issue with -suest- in Version 11 with the new
> factor coding.  Basically, I get the error "too many base levels
> specified" when I try to "suest" models that share the same factor
> variable, but include different values of the factor variable. When I
> revert to the old "xi" syntax, the models can be combined with -suest-
> as expected.  I can recreate this problem with the following example:
>
> sysuse auto.dta
> xtile pricecat=price, nq(4)
> logistic  foreign trunk headroom i.pricecat if pricecat<=2
> est store m1
> logistic  foreign trunk headroom  mpg  i.pricecat if pricecat>2
> est store m2
> suest m1 m2
> *error produced "too many base levels specified"
> xi: logistic  foreign trunk headroom i.pricecat if pricecat<=2
> est store m3
> xi: logistic  foreign trunk headroom mpg i.pricecat if pricecat>2
> est store m4
> suest m3 m4
> *suest produces following results:
>
> Simultaneous results for m3, m4
>
>                                                    Number of obs   =         74
>
> ------------------------------------------------------------------------------
>               |               Robust
>               |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
> -------------+----------------------------------------------------------------
> m3_foreign   |
>         trunk |  -.1917506   .2408272    -0.80   0.426    -.6637633     .280262
>      headroom |   .1620987    .842212     0.19   0.847    -1.488607    1.812804
> _Ipricecat_2 |   .3446382   .7647411     0.45   0.652    -1.154227    1.843503
> _Ipricecat_3 |  (omitted)
> _Ipricecat_4 |  (omitted)
>         _cons |   .2902781   1.124931     0.26   0.796    -1.914546    2.495102
> -------------+----------------------------------------------------------------
> m4_foreign   |
>         trunk |  -.5303469   .2455778    -2.16   0.031     -1.01167   -.0490233
>      headroom |   .2068252   .7818994     0.26   0.791    -1.325669     1.73932
>           mpg |   .1186309   .0674696     1.76   0.079    -.0136071    .2508688
> _Ipricecat_2 |  (omitted)
> _Ipricecat_3 |  -.4138492   1.126895    -0.37   0.713    -2.622524    1.794825
> _Ipricecat_4 |  (omitted)
>         _cons |   4.712297   4.920069     0.96   0.338    -4.930861    14.35545
> ------------------------------------------------------------------------------
>
> Anyone have an idea why -suest- cannot combine models with the new
> factor variable coding? Am I making an error, or cound this be
> intentional? Otherwise is it a bug?
>
> Many thanks. Tim
> *
> *   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/