Re: st: -warning: variables in at() list not found in model after mfx or mfx2

[Richard Williams <Richard.A.Williams.5@ND.edu>]

At 05:56 PM 7/20/2007, Jean Marie Linhart, StataCorp. wrote:
> Mirko <mirko.moro at gmail dot com> and others have reported a mysterious
> warning message with -mfx2- after -regoprob-, identifying that it is
> generated by -mfx- when the variable list in the original estimation and
> -at()- specification are long.  We have been able to reproduce this
> problem.
>
> We believe the results from -mfx- are correct; the warning message is not
> correct and can be ignored.  We want to do a little more exploring, 
> and we will
> report back to the list with a definitive answer.
Thanks Jean.  -mfx- subprograms -at- and -Parseatlist- include a lot 
of equals signs when working with strings, and also use the string 
functions -trim- and -index-.  I think some or all of these can have 
problems once strings get to be more than 244 characters.  In 
particular, mfx subprogram Parseatlist includes the lines

                if index("`vars'","`thing1'")==0 {
                        local unfnd ="`unfnd'"+ " `thing1'"
                }

That eventually leads to a false warning message if thing1 does not 
occur within the first 244 characters of vars.  I changed those lines to

                local unfnd: list thing1 - vars

and the error went away.  Anyway, I think that is among the changes 
that could/should be made in mfx.

As far as I can tell though, you are correct in saying that, despite 
the incorrect warning, mfx does produce the correct results.




-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW:    http://www.nd.edu/~rwilliam

*
*   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/