I'm trying to adapt some S-Plus scripts to run in R (1.2.2, Windows). In
one of these scripts, I've bootstrapped the prediction success rate under
the discriminant function (lda). The bootstrap() functions are proprietary
to S-Plus and there aren't exact equivalents in R. The closest is
Canty's
library boot based on the Davidson and Hinkley book. Unfortunately, I
haven't been able to map the commands I used in the bootstrap() function of
S-Plus to the boot() function in the boot library of R (and also S-Plus).
Below is an example of one line from this S-Plus script. The particular
element I'm having trouble mapping to boot() is the
group argument. This makes it pretty straightforward to work with
functions that operate on multiple groups. boot() has a strata argument,
which *seems* like it ought to be the rough equivalent of group, but its
operation is not entirely transparent. Can someone point me to an example
where boot() is used as I've illustrated below with S-Plus' bootstrap()
function? Unless I've missed something, there isn't an illustration in
VR3
or in the help for boot. (There is a bivariate example, but the particular
example doesn't translate well to something like a discriminant analysis).
Thanks for pointers.
boot.lda.success<-bootstrap(apehum, hitratio(apehum[,1],
predict(lda(apehum[,2:11], apehum[,1]))$class)$error, group=apehum[,1],
trace=F, B=250)
====================Dr. Marc R. Feldesman
Professor and Chairman
Anthropology Department
Portland State University
1721 SW Broadway
Portland, Oregon 97201
email: feldesmanm at pdx.edu
phone: 503-725-3081
fax: 503-725-3905
http://web.pdx.edu/~h1mf
PGP Key Available On Request
=====================
"Anyway, no drug, not even alcohol, causes the fundamental ills of society.
If we're looking for the source of our troubles, we shouldn't test
people
for drugs, we should test them for stupidity, ignorance, greed and love of
power." P.J. O'Rourke
Powered by Optiplochoerus and Windows 2000 (scary isn't it?)
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
https://stat.ethz.ch/pipermail/r-help/attachments/20010316/e0f776d3/attachment.html
