Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: LaTeX, Stata and mass production


From   Ian Watson <i.watson@econ.usyd.edu.au>
To   statalist <statalist@hsphsun2.harvard.edu>
Subject   st: LaTeX, Stata and mass production
Date   Thu, 21 Oct 2004 19:23:53 +1000

Dave Moore recently wrote, in reponse to a long thread about nice
tables from Stata (part of the Wish List thread):

      Second, I must be the only one who thinks the
      output generated by stat packages (including SPSS, SAS,
      Stata, LIMDEP, Systat, Gauss, Splus, etc.) is not suitable 
      for reports, because I've never had any desire to copy
      output directly from an output window to a word
      processor to use as is. At best, copying from one to 
      the other saves having to type the relevant information
      manually, but I typically have to spend a great deal of
      time reformatting everything to make it pretty enough
      for my purposes. In fact, I find utilities such as -outreg- 
      infinitely better suited to this task. In fact, in an
      environment where generating reams of tables is
      routine, the thought of copying and pasting hundreds of
      tables is particularly off-putting. Like Marcella, I
      basically favor linking the desired output to
      Excel/Word in such a way that the formatting is
      essentially automatic. I write programs to generate output
      files that I load into Excel (creating Stata data sets
      seems an unnecessary step), which makes the whole
      process faster than copy-paste-reformat,
      copy-paste-reformat, ....
  
I would agree wholeheartedly.

As someone who has to also mass produce descriptive tables and make
them look nice, I long ago abandoned the copy-paste-reformat routine.
When I used SPSS there was no easy way to automate it. On the other
hand, a little bit of Stata programming allows one to automate all
these mass production jobs and run a series of batch files. A sudden
decision to change the weighting, or some other hiccough, simply
results in a new batch run (perhaps 10 minutes which you have a cup of
tea). With the copy-paste-reformat routine, you're looking at days of
work ...

While there a various options of getting the tables into Word/Excel,
an option which is truly automatic is to go the LaTeX route. The
choice depends on whether it's essential to have a Word version of the
final "nice" presentation or not. If the end user / audience only
requires the final "nice" presentation as a PDF file, then the LaTeX
route produces beautiful output straight from Stata. (More on this in
a moment). If they require a Word version (but not necessarily as
pretty as the PDF), then the PDF file can be converted into Word.
Tools for this include OmniPage Pro and ScanSoft PDF Converter (the
latter is considerably cheaper than the former). The downside is the
need to learn LaTeX ...

Finally, I come to the point of my email. I'm currently revising two
Stata programs I wrote a couple of years ago -latab- and -latabstat-
which turn the output from tabulate and tabstat into LaTeX code. I've
found that the approach I took in LaTeX terms doesn't really suit me
anymore, and I'm considering other options (mainly to do with the tops
and tails of the tables). I really like Ben Jann's wonderful -estout-
which uses an innovative approach to user-defined output, and am
considering variations on that. Anyway, I'd like to hear from LaTeX
users (off list), and particularly anyone who's used -latab- or
-latabstat-, about what improvements to the LaTeX approach to tables
they might find useful.

I'm also developing a new Stata-LaTeX program - called "bigtable"
(btab.ado) - for building large tables which incorporate many vertical
variables cross-tabbed by a single horizontal variable. Row or column
percentages, control over decimal points, bolding for subheadings
etc. I'm also considering allowing user-defined delimiters which would
allow for HTML output as well as LaTeX. My motivation has been the
need to produce these kinds of tables in a mass production way for use
in the reports which our centre regularly produces. I'd be keen to
hear from anyone (off list) with an interest in beta testing
"bigtable" and suggesting improvements.


-- 
Kind regards,
 Ian                          

=================================
Ian Watson
Senior Researcher
acirrt
University of Sydney
NSW 2006
Australia

Phone: 02 9351 5622
  Fax: 02 9351 5615

Email: i.watson@econ.usyd.edu.au
  Web: www.acirrt.com
================================

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index