Re: st: Use of "Goto" command in Mata

[Nick Cox <njcoxstata@gmail.com>]

Another way to do it is

forval i = 1/4 {
         use Results, clear
         keep if id == `i'

         <call up Mata-based code>
}

On Mon, Jan 2, 2012 at 11:04 AM, Ozgur Ererdem <ozgur.ererdem@ecopa.com> wrote:
> Thank you very much Nick.
>
> I will check the use of select() as you recommended.
>
> I need Mata for this operation. I wrote down a very simplified version of the code in my previous message. Actually what I am trying to do is to solve non-linear equation systems for each block of observation (about 7000).
>
> Thank you again.
>
> Ozgur
>
>
>
>
> It is not clear that you need Mata at all here. But the problem arises because "end" is ambiguous. So, you can't do it this way.
>
> Assuming that there is a case for Mata  I would read all the data into Mata and loop over 1..4 and use -select()- to select blocks of observations. -goto- could probably be made to work but is not needed when you know you want to loop over 1..4.
>
> Nick
>
> On Mon, Jan 2, 2012 at 10:02 AM, Ozgur Ererdem <ozgur.ererdem@ecopa.com> wrote:
>
>> I am trying to run a loop within Mata by calling, at each loop, a different part of Stata results, distinguished by a variable named "id". For example, during the first loop, only results with id==1 will be used, then the ones with id==2 etc.
>>
>> I thought that the most adapted way to do it was to use the old-style "goto" command. Here's what I wrote:
>>
>> clear mata
>> mata
>> id = st_data(.,"id")
>> y=0
>> B1: y=y+1
>> mata stata use "Results.dta",clear
>> st_numscalar("y", y)
>> mata stata keep if id==y
>> beta = st_data(., "beta")
>> beta
>> y
>> if (y<4) goto B1
>> end
>>
>> When I do this, the loop does not start. Thus I obtain only results for y==1. The error I get is the following:
>>
>> : if (y<4) goto B1
>>> end
>> unexpected end of line
>>
>> I also tried to run the loop by typing "if (y<4) {} goto B1" but I get the following message:
>>
>> system error:  UDpaddr
>> r(3000);

*
*   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/