Re: st: RE: Metan help

[Sripal Kumar <sripalkumar@gmail.com>]

Thanks Nick.  I figured out that it was not my Mac but the latest
version of metan which does not support a dialog box.  Similarly, I
guess there were some modification to the metabias command and the
command I used to use previously did not work.  Now it does.
thanks,
Sripal.

On Mon, Mar 22, 2010 at 7:46 AM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
> I don't understand what you mean in your reference to profile.do. Nor can I, or as far as I can guess anybody else, explain why you get an error with -metabias-. You should explain what error you mean.
>
> But the rest looks relatively simple and nothing to do with your use of a Mac.
>
> You can't use -metan- through a dialog if you don't have a dialog file installed, and that's the problem, as the error message you quote does include.
>
> -metan- doesn't qualify as a programming command, but it is a user-written command. The latest version of -metan- appears to be at
>
> SJ-9-2  sbe24_3 . . . . . . . . metan: fixed- and random-effects meta-analysis
>        . . . . . . . . . Harris, Bradburn, Deeks, Harbord, Altman, and Sterne
>        (help funnel, labbe, metan if installed)
>        Q2/09   SJ 9(2):327
>        by() variables were being converted to string format and
>        the predictive interval option rfdist had an error when
>        used for nonratio measures; these problems are fixed
>
> and the files for this version do not include a dialog file. Your installation, among other things, copied from that source.
>
> Thomas Steichen independently maintains a -metadialog- package on SSC, which may be what you want. I can't comment on whether the dialog files there are up-to-date for -metan-. Tom or others should be able to comment further.
>
> Nick
> n.j.cox@durham.ac.uk
>
> Sripal Kumar
>
> I use Stata 10 for Mac.  Previously when I used to use it with
> windows, it used to work well.  I have the following problem on my
> mac.  I installed the meta analysis commands from
>
> . net from http://www.stata-press.com/data/mais
>
> I have a user written command saved as profile.do for meta-analysis.
>
> When I use the user written command menu:  User-->meta analysis-->
> metan, I get the "Dialog for metan not found:  Either metan is a
> programming command or a user-written command for which no dialog box
> has been written"  The same thing happens when I type in "db metan".
> However, I still can run the command if I use "metan variable1
> variable 2 OR etc".
>
> Is this something to do with MAC?
>
> I get an error for metabias as well though I am not using the dialog
> box command.
>
> *
> *   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/
>

*
*   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/