Davide,
the problem is in a part of your code that you are not showing. Your
program, so far, does not use evaluation (the evaluating equal sign).
See also here: http://statajournal.com/article.html?article=pr0045 .
As soon as you manipulate `regressors' with an equal sign, the limit
bites. Without it, you should be fine, as the following example shows.
Eva
************************************************
capture program drop mycommand
program mycommand
syntax namelist [if] [in] [aweight fweight pweight] [, *]
gettoken command 0 : namelist
gettoken depvar regressors : 0
`command' `depvar' `regressors' `if' [`weight' `exp'] , `options'
end
* create 150 regressors
clear
set obs 5000
set seed 123
forvalues x = 1/150 {
gen x`x' = invnorm(uniform())
}
gen eps = invnorm(uniform())
gen y = 5
forvalues x = 1/150 {
replace y = y + `x'*x`x'
}
replace y = y + eps
* use regress
regress y x*
* use mycommand
qui ds x*
mycommand regress y `r(varlist)'
************************************************
2009/3/16 Davide Cantoni <[email protected]>:
> Dear statalisters,
>
> I have a question related to programming a new program/ado file. I am
> encountering the problem that one of the arguments of my program might
> be too long as a string, going beyond the limit of 244 characters.
> Specifically, my code looks like this
>
> *** begin code
>
> program mycommand
> syntax namelist [if] [in] [aweight fweight pweight], ...
>
> gettoken command 0 : namelist
> gettoken depvar regressors : 0
>
> `command' `depvar' `regressors' `if' [`weight' `exp'], ...
>
> ...
>
> end
>
> *** end code
>
> (to explain: the idea is to have a two-word command, where the second
> word is a regression command built-in in Stata. So, for example, one
> would run "mycommand areg y x1 x2 x3..." and then -mycommand- would
> use -areg- within its routine)
>
> I am encountering problems when running -mycommand- with a list of
> regressors that is longer than 244 characters. Then, the list is
> arbitrarily truncated after that limit, which of course can create
> problems. What would a savvy programmer to do avoid this problem? I
> cannot imagine that any list of regressors has to be shorter than 244.
>
> Thanks for your help and suggestions,
>
> Davide
*
* 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/
