[R] leaps and bounds

Prof Brian D Ripley ripley at stats.ox.ac.uk
Sun Aug 15 08:52:29 CEST 1999

On Sat, 14 Aug 1999, Troels Ring wrote:

> Dear friends. 

> On the Bayesian averaging homepage 
> http://www.research.att.com/~volinsky/bma.html I found 
> some S code some of which perhaps may run in R. 
> There was a call to an algorithm possibly within S but not 
> supported by R 64.1: "leaps and bounds". I guess it is a minimization
> step. 
> Can anyone clarify the algorithm and perhaps even give a pointer to some
code ?
> I guess this may be somewhat above my level of understanding 
> but I feel much attracted to the concept of model averaging, 
> as far as I can grasp it, and would like to experiment with it.

That page has a rather individual (and I think very limited) view of
Bayesian model averaging. In particular, I think you are looking at links
to Raftery's work which has been on statlib for some time, afunction called
bicreg() that calls leaps(), an S version 2 (Blue book) function. A version
of leaps is available in package leaps on CRAN, and may (I have not tried
it) enable you to use bicreg.

However, as leaps is an algorithm to select the best subset of size k, it
goes directly against my understanding of model averaging, which involves
averaging over all subsets and hence in the linear regression context is a
form of downweighting of regression coefficients. 

There are many other Bayesians involved with model averaging with
well-known papers not mentioned on that site. In particular, I would
suggest you look at David Draper's work (University of Bath: has S software
and a draft book, which was available when I last looked) and for
mainstream Bayesian approaches to regression see the 1995 book Bayesian
Data Analysis by Gelman, Carlin, Stern and Rubin.  To be doing model
averaging in the Raftery sense needs a very particular prior over the
regression coeficients that few other Bayesian I know find credible.

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list