Hi everyone,
I try making decision trees and random forest using the packages rpart and
party. I'm already stuck at t he first step. Each time when I enter the code
either 1. R takes more than an hour. I haven't waited long enough to see if
there's a result but it doesn't look like it! When I hit the "stop" button it
also freezes and I need to force quit R. This happens with rpart().
2. Or I get the message :
Error: cannot allocate vector of size 5.0 GbIn addition: Warning messages:1: In
cbind(RET, tr[[i]]) : Reached total allocation of 16287Mb: see
help(memory.size)2: In cbind(RET, tr[[i]]) : Reached total allocation of
16287Mb: see help(memory.size)3: In cbind(RET, tr[[i]]) : Reached total
allocation of 16287Mb: see help(memory.size)4: In cbind(RET, tr[[i]]) :
Reached total allocation of 16287Mb: see help(memory.size)
When I look at Windows task manager it goes from "in use: 4GB" to use "in use:
14.5 GB" causing it to have no memory left (15.9 GB is the limit on my
computer). The trainset is big (almost 9 million records and 13 variables). I
already increased the memory.limit() but it didn't work. This happens with
ctree(), cforest().
I don't have much technical knowledge and I am a beginner at R. I use the
64-bit version on Windows.
Examples of the code that I used:dt <- rpart(Product ~ Age + TotalChildren +
NumberCarsOwned, data=TrainData, method="class",
control=rpart.control(minsplit=50, cp=0, xval=0))
Formula <- Product ~ Age + TotalChildren + NumberCarsOwnedctree <-
ctree(Formula, data=TrainData)
rf <- cforest(rFormula, data=TrainData)
On a smaller data set (of 18.000 records) it does seem to work...How can I make
it work on my dataset? Is there something in the arguments that I should
change.
Kind regards,
Kim
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
