Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: -cmp- does not work properly - wrong installation?


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: -cmp- does not work properly - wrong installation?
Date   Fri, 20 Apr 2012 19:26:58 +0100

The remaining problems are with your data. You need to look at
-correlate- or -findit collinear- to find user-written add-ons.

-nrtolerance()- is an option of -ml- that you can set from -cmp-. You
can find this in the help for -cmp- as it is documented. Use the
search facility in the Viewer. Click through to help and documentation
for -ml-.

Nick

On Fri, Apr 20, 2012 at 7:18 PM, Laura R. <laura.roh@googlemail.com> wrote:
> Part of the problem is solved! Great, thanks alot!
>
> I updated my Stata 11.1 version, as Nick has just suggested, and then
> (just in case) repeated the other commands everyone suggested (but it
> looked as they were not neccessary, though).
>
> Now a large part of the error messages has disappeared, and the -cmp-
> model with 1 or 2 equations run all the way through. (I realised now
> that they did not do it before).
>
> The do file with the publicly available data I said I tried runs fine,
> and I get results without any error messages.
>
> The only error message left when using my data is
>
> *---------------------------- error message begin --------------------------
>
> Warning: regressor matrix for y1 equation appears ill-conditioned.
> (Condition number = 70.887255.)
> This might prevent convergence. If it does, and if you have not done
> so already, you may need to remove nearly
> collinear regressors to achieve convergence. Or you may need to add a
> nrtolerance(#) or nonrtolerance option to the comma
>> nd line.
> See cmp tips for more information.
>
> *------------------------ error  message end -----------------------------
>
> for both y1 and y2.
>
> y1 is the number of children, and I did ordered probit, but I get the
> same error message when using the "continuous" specification in the
> -cmp- model.
>
> y2 is clearly bivariate (0 or 1), and I used probit specification in
> the -cmp- command.
>
> Do you have ideas for reasons for this error message and how to solve it?
>
> I mean, it converges, but it's not nice to have this error message. I
> also don't know which independent variables might be nearly collinear,
> as I receive the error message also with few indep. variables which
> are clearly different from each other.
>
> Also, do you know what this "nrtolerance(#) or nonrtolerance option "
> exactly is, and how you use it? I have read everything in the -cmp-
> help file, but I did not get it so far ...
>

*
*   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/


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