Dear Stas, thanks for getting back to me with that. I have a query
regarding mlogit, hope you don't mind me asking...Do you know why Stata
gives an error message "too few categories" when you run mlogit with
frequency weights, but if you run the same mlogit model without specifying
frequency weights then it doesn't give this error message, even if the
dependent variable is the same in both (0 or 1 values for depvar)?
Cheers
Claire
Quoting Stas Kolenikov <[email protected]>:
> --- In statalist, Claire Johnstone wrote:
> > Dear Stata-users, I have 2 queries re. Conditional Logit (I've searched
> > online help & don't have a manual). These are:
> > 1. Is there a maximum number of alternatives you can have in a choice
> > set?
>
> As long as -clogit- is the same as -xtlogit-, the question can be restated
> as "What is the maximum number of observations per panel?", and you can
> find that in -help limits- to be the maximum matsize (800 for Intercooled,
> 11K for SE).
>
> > 2. Can you have multiple positive responses to alternatives in a single
> > choice set?
>
> Again, as long as you can have multiple 1's in the panel setting, this
> should work for the regular -clogit-, too. By trial and error,
>
> . clogit for pri, group(rep)
>
> note: multiple positive outcomes within groups encountered.
> note: 2 groups (10 obs) dropped due to all positive or
> all negative outcomes.
> Iteration 0: log likelihood = -23.108062
> Iteration 1: log likelihood = -23.107154
>
> Conditional (fixed-effects) logistic regression Number of obs =
> 59
> LR chi2(1) =
> 0.00
> Prob > chi2 =
> 0.9660
> Log likelihood = -23.107154 Pseudo R2 =
> 0.0000
>
>
----------------------------------------------------------------------------
--
> foreign | Coef. Std. Err. z P>|z| [95% Conf.
> Interval]
> -------------
+----------------------------------------------------------------
> price | 5.58e-06 .0001307 0.04 0.966 -.0002507
> .0002618
>
----------------------------------------------------------------------------
--
>
> all Stata does is it notes to you that there are multiple positive
> outcomes, but does not halt on that.
>
> As for the weights question, I don't think there is any way around it
> except for explicitly -expand-ing your data, so if you really have tons of
> implied observations under [fw], you might be in trouble as regarding the
> number of observations per panel / choice set.
>
> --- Stas Kolenikov
> -- Ph.D. student in Statistics at UNC-Chapel Hill
> - http://www.komkon.org/~tacik/ -- [email protected]
>
> * This e-mail and all attachments to it are not intended to provide any
> * reasonable point of view and was transmitted to you in error. It
> * should be immediately deleted by all recipients unless they really
> * enjoy communicating with the author :). Other restrictions apply.
>
> *
> * 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/
>
--
*
* 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/
