Message-ID: <20030514113700.51782.qmail@web14202.mail.yahoo.com>
Date: 2003-05-14T11:37:00Z
From: John Hendrickx
Subject: mcl models, percentages
I've put two packages for R on my home page at
http://www.xs4all.nl/~jhckx/R/. The "pcnt" package is for multiway
percentage tables. I've posted a first effort called "ctab" on this
group and a request for enhancing "ftable" with percentages on the
wishlist.
The "mcl" package is for estimating multinomial logistic models using
conditional logistic regression. This gives greater flexibility in
imposing restrictions on the dependent variable. One application is
to estimate loglinear models for sqaure tables (e.g.
quasi-independence, quasi-symmetry) with covariates. The mcl package
therefore contains a number of functions for models for square tables
as well.
A caveat is that "clogit" in R doesn't produce the same estimates as
"multilog", although the likelihood functions for both models are the
same. The maximum absolute difference is 0.0034, the mean absolute
difference is 0.00069. Stata's "clogit" and "mlogit" produce the same
estimates and match those of "multilog" to at least 6 decimal points
accuracy. See the notes in http://www.xs4all.nl/~jhckx/R/mcl.html Can
anyone shed any light on this?
John Hendrickx