Re: st: xtlogit: panel data transformation's recast to double makes model incomputable

[Tom <tommedema@gmail.com>]

Dear Maarten,

Apologies for the wrong naming of Stata, I read the FAQ all the way up
to this point.

How is Nick's suggestion a solution when the objective is to figure
out the significance over the other independent variables on my
dependent variable? Merely using two variables is not going to give me
any useful information.

Insofar noone has explained why the computation is unsuccessful. It'd
be especially helpful if someone could explain why less precision
would allow the maximizer to do it's job.

Kind regards,
Tom

On Tue, Apr 2, 2013 at 2:00 PM, Maarten Buis <maartenlbuis@gmail.com> wrote:
> On Tue, Apr 2, 2013 at 1:38 PM, Tom  wrote:
>> I am experiencing a very peculiar problem in STATA. I have a panel
>> dataset with 8000 groups and 4 million observations. I want to run a
>> fixed effects panel binary logistic regression (xtlogit).
>>
>> As you can read in [my original question][1], STATA's maximization
>> algorithm cannot compute a solution when using `xtlogit`. I have
>> reduced this problem to occurring when transforming the data to panel
>> data, because:
>>
>> 1. running `logit depvar indepvars` works just fine
>> 2. `logit depvar indepvars` does not maximize when first running
>> `xtdata indepvars, fe clear`
>>
>> Therefore `xtdata` does something that causes the maximizer to fail.
>
> I don't know how things work in STATA but if you mean Stata(*), then I
> do not think that that is the right diagnosis. -xtlogit- does not do
> what -xtdata- does, so the precision of the transformed variables is
> not the source of your problem, since -xtlogit- does not transform the
> variables. Anyhow, a "solution" that rests on less precision during
> compution would be extremely suspicious. You don't like it, but the
> solution remains the one that Nick proposed earlier: simplify your
> model.
>
> -- Maarten
>
> (*) A quote from the Statalist FAQ, which everybody is asked to read
> _before_ posting to Statalist:
>
> Stata is an invented word, not an acronym, and should not appear with
> all letters capitalized: please write "Stata", not "STATA".
>
> ---------------------------------
> Maarten L. Buis
> WZB
> Reichpietschufer 50
> 10785 Berlin
> Germany
>
> http://www.maartenbuis.nl
> ---------------------------------
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/