--- On Tue, 16/6/09, timothy adler wrote:
> I am just surprised that this fractional logit and
> compositional data analysis appears in stata, but the
> econometrics behind it is underexplored or not (fully)
> discussed.
Notice that what appeared in Stata is mostly user written
with the exception of the way we can trick -glm- to
fit a fractional logit model. StataCorp is for good
reason conservative whith what it implements in official
Stata, this way they won't have to retract any commands.
However, StataCorp allows and encourages users to write
their own commands implementing whatever technique a
user may think is useful, regardless whether the
technique has fully matured or is still "bleeding edge".
My impression of the state of the art with respect to
compositional data analysis ist that this may be
underexplored within econometrics (though Jeff
Wooldridge has definately worked and published on it),
but in other sciences real progress is being made, in
particular within geography. There are enough open
questions to justify it not being implemented in
official Stata (e.g. how to handle proportions of
exactly 0 or 1), but that is more a matter that a
concensus needs to appear of what reasonable but
faulty solution is the most reasonable/least faulty.
-- Maarten
Another reference is:
A Buccianti, G Mateu-Figueras and V Pawlowsky-Glahn (2006)
Compositional Data Analysis in the Geosciences: From Theory
to Practice. Geological Society special publication, no. 264
London : Geological Society.
-----------------------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* 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/
