On May 17, 2010, at 1:16 PM, Michael Norman Mitchell wrote:
Thanks for posting this to bring this to our (and my) attention. I
always believe very strongly in using this kind of "weaving"
approach and find it very efficient. Are you (or is anyone on the
Statalist) aware of a tool similar to this that weaves comments,
commands, output, and graphs into a Word document? At my work, we
are in constant need of creating such integrated outputs and the
standard format is a Word document.
I'm afraid your options here are going to be a lot more restricted if
your target format is Word, since the people who have worked on this
problem have tended to prefer LaTeX-based approaches or similar. The
one MS Office-based product I'm aware of for doing reproducible
research is inference, from Blue Reference (http://inference.us/).
Unfortunately, I don't believe they currently support Stata (I could
be wrong here, since I haven't looked at this in a while). I don't
use Word myself, so there may be other MS Office-based options of
which I am unaware. One could certainly write something that would
work with Office (as inference demonstrates), however you'd have to
use VSTO and you'd be stuck with MS's upgrade cycles.
Another alternative is to use one of the other tools for reproducible
research (e.g., StatWeave) and then translate the final product into
Word. This is the strategy I tend to favor when I have to deliver a
final product in Word format, and there are several options here
(though none of them is perfect). For example, you can translate from
LaTeX to RTF using latex2rtf, and then open the RTF file in Word (this
is what I normally use). Another option is to use TeX4ht to translate
from LaTeX into Open Office, and then use Open Office to save the
file(s) in Word format. There may be better tool-chains for doing
this out there -- I just don't do it often enough to go looking.
Ideally, you'd get people to covert to PDF as the target format (after
all, you can work with PDFs on all platforms), and then you'd be in