Martin--
It's not necessary to reduce to a two-dimensional problem; you can use
nested loops over values of many column and "supercolumn" variables,
loop over distinct values of the row variable, and build up b and V
matrices to post as estimates (one per column). If you have more than
2 statistics per cell to tabulate, or any strangeness in the table
layout, it will usually be easier to write the values to a TeX file
yourself with -file- than to trick -estout- or another program into
displaying "estimates" though...
On Fri, Mar 20, 2009 at 3:23 PM, Martin Weiss <[email protected]> wrote:
> <>
>
> Those old files that somehow made their way into your PERSONAL folder and never got killed... Once they are gone, -adoupdate- takes care of the update in no time. Sorry for all the fuss and thanks to Austin for pointing me in the right direction.
>
> Importantly, though, my intitial suspicion that one needs to reduce the problem to a two-dimensional one via -egen, group()- seems to be corroborated?
>
> HTH
> Martin
*
* 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/