On 3/2/09, John Sandberg, Prof. <[email protected]> wrote:
> Thanks for the input about bs4rw. Stats Can does have a preference for
> SAS, and from what I know of how analysts have approached this, there is
> an ad-hoc approach using a number of software packages depending on the
> specific analysis.
A part of that is a production issue. From what I've seen over the
last few years at user talks, one can twist the hands of Stata to
produce any sort of text-looking output you need (and for my latest
paper, I had way more -file- commands with LaTeX formatting of
-estimates- and -return- results than the lines with statistical
commands... but again I am the kind of guy who re-writes stuff from
JASA or Annals in Mata when needed for my research purposes :)) ). SAS
on the other hand spits HTML tables by default -- for busy people,
that's a perfect solution. If you need to produce a report with 300
tables, and you need to do ten reports like that a month, you would
want to have something basic and standard, like -html- or -rtf- or
-xml- option of -table-/-tabulate-. Stata Corp said that they are
looking at fancyfying output very seriously, tracking what's going on
at SSC and all the output formatting routines available there.
Simplifying production might improve appeal of Stata to the official
statistics audience.
--
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
* 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/
