Re: st: Variable number of matrixes as arguments

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

At 06:01 PM 7/26/2005 -0400, Herb Smith wrote:

> But what if I want the routine to return, on occasion, some values
> calculated over these matrixes such as (for simplicity!) the number
> of non-zero cells in the new matrix, B?  Assume that this number is
> being calculated within the -gprvector- routine in the local `cells':
>
> How do I return this result along with the new matrix, B?  I tried (by
> analogy)
>
>         syntax, input(name) output(name) [cells(name)]
>
> and then executed
>
>         gprvector, input(A) output(B) cells(z)
>
> but the follow-up command
>
>         dis z
>
> draws a blank.
This is hard to answer without seeing more code and testing it out, but 
following are my guesses on this.

First off, it is always possible z never got assigned anything, so you 
could try -set trace on- to make sure your program is doing what you want 
it to.

Second, within the program, what does the code for `cells' look like, i.e. 
how are you assigning it a value?  I think it should be something like

scalar `cells' = some number or another

My guess (quite possibly wrong) is that you are using local commands 
instead, e.g. maybe you have

local `cells' = some number or another

which expands to

local z = some number or another

and `z' (which is not the same as z) disappears after the routine is run.

But if that isn't it -- It is a little lazy, but if I am having trouble 
getting my routines to communicate with each other I use global 
variables.  So, within -gprvector-, have something like

global z = some number or another

After having executed -gprvector-, you can give the command

display $z

I suspect the most common way to get what you want is to set your program 
up as rclass, sclass or eclass.  So the program might look something like

program define gprvector, rclass
        ... do bunch of stuff
        local z = some number or another
        ... more stuff
        return local cells `z'
        ... finish program

Back to main program or interactive run:

display r(cells)

-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX:    (574)288-4373
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW (personal):    http://www.nd.edu/~rwilliam
WWW (department):    http://www.nd.edu/~soc

*
*   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/