The authors of -movestay- do not appear
to be members of Statalist.
I think Renzo is right. His example is
fixed in the following way. Copy the program
to a new file. After the block of code
tokenize "`select'", parse("="); // define vars for regression equation
if ("`2'"!="=") {;
tokenize "`select'";
local y_prob `1';
macro shift;
local x_prob `*';
};
else {;
local y_prob `1';
local x_prob `3';
};
a new line is needed to unpack wildcards:
unab x_prob : `x_prob' ;
My hunch is that this should not prove problematic
in other ways.
Nick
[email protected]
Renzo Comolli
>
> I want to thank the authors of -movestay-. It is an important
> contribution,
> and it will help me a lot.
>
> I think there might be an issue with varlist abbreviations though.
> Sometimes, abbreviating a varlist with an * causes the error message
> "initial values not feasible".
> It does not happen with the dataset the authors use for their
> example, so I
> posted my own (admittedly contrived) dataset.
>
> E.g.
> . use http://www.econ.yale.edu/~rc255/data/for_question_movestay.dta
> . movestay income age, select(group race_2 race_3) // this
> works as intended
> . movestay income age, select(group race_*)
>
> The first instruction works as expected;
> the second one does not (albeit in this dataset race_*
> corresponds to race_2
> race_3)
*
* 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/
