Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: collinear problem with asclogit


From   rgates@stata.com (Richard Gates)
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: collinear problem with asclogit
Date   Wed, 06 Feb 2008 08:27:50 -0600

Erly <erlyana.c@gmail.com> is observing the following behavior from
-asclogit-.
 
> asclogit use km2kd price, case(pidlink) alternatives(mode)
> casevars(age male married hs1 hs2 educ2 educ3 educ4 expmcapth region2
> region3 hhsize employment adlindex serill) nolog collinear
> note: model has collinear variables; convergence may not be achieved
> convergence not achieved
> 
> Alternative-specific conditional logit         Number of obs      =       2264
> Case variable: pidlink                         Number of cases    =        566
> 
> Alternative variable: mode                     Alts per case: min =          4
>                                                               avg =        4.0
>                                                               max =          4
> 
>                                                   Wald chi2(44)   =          .
> Log likelihood = -368.13846                       Prob > chi2     =          .
> 
> ------------------------------------------------------------------------------
>          use |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
> -------------+----------------------------------------------------------------
> convergence not achieved
> 

This looks like a display bug.  If my counting is correct, the table should
have 50 coefficents (2 a.s.  coefficients, 45 c.s., and 3 constant terms) and
at least 3 coefficents (47-44) would have missing values, but the display ado
code is terminating early.  I do not see anything obvious in the display code
that would cause this, nor can I recreate it.  Perhaps something did not get
cached properly in -ereturn-.  If Erly could send me the data and a -do- file
that recreates the problem it would make it easier investigate.

-Rich
rgates@stata.com
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index