[R] using randomForest() with matrix() as input results to an Error: protect(): protection stack overflow

Nick Fikas elvenpath666 at gmail.com
Wed Jan 31 16:56:45 CET 2018


I'm trying to figure out a solution online but couldn't so far. I realized
that others dealt with such an error but their suggestions didn't work for

So, I'm trying to run the randomForest() using this command:

rf = randomForest(classes~., data=as.matrix(train), mtry=5, ntree=2000,

and it always results in *Error: protect(): protection stack overflow.*

As you can see, I have already turned the training dataset into a matrix
but this didn't fix anything.

> Cstack_info()

      size    current  direction eval_depth
   7969177      13104          1          2

I also tried the options(expressions = 12e4) but nothing really changed.

The PC I'm running this on has 12 GB RAM and runs on Linux.

The dim(train) returns: 50 20040.

Is there something else to try, or I should run it on a different PC?


	[[alternative HTML version deleted]]

More information about the R-help mailing list