R help - Feb 2011 - e1071's Naive Bayes with Weighted Data

Hello fellow R programmers,

I'm trying to use package e1071's naiveBayes function to create a model
with
weighted data.  See example below, variable "d" is a count variable
that
provides the # of records for the given observation combination.  Is anyone
aware of a "weight" argument to this method?  I've been
unsuccessful in my
research.

Thanks,

Mike

library(e1071)

a<-c(1,1,1,1,0,0,0,0)
b<-c("A","B","A","B","A","B","A","B")
c<-c("X","X","Y","Y","X","X","Y","Y")
d<-c(4,3,2,1,20,15,11,4)

input<-data.frame(a,b,c,d)

## NAIVE BAYES - NO WEIGHT ##

model <- naiveBayes(a ~ factor(b)+factor(c), data = input)


## NAIVE BAYES - WEIGHT by d? ##
model <-naiveBayers(a~factor(b)+factor(c),data=input,weight=d)



-- 
Michael Schumacher
mike.schumacher@gmail.com
Manager, Data & Analytics
ValueClick

	[[alternative HTML version deleted]]