Random Forest prediction questions
From: Dror
Hi, I need help with the randomForest prediction. i run the folowing code:
iris.rf <- randomForest(Species ~ ., data=iris, importance=TRUE,keep.forest=TRUE, proximity=TRUE) pr<-predict(iris.rf,iris,predict.all=T) iris.rf$votes[53,]
setosa versicolor virginica 0.0000000 0.8074866 0.1925134
table(pr$individual[53,])/500
versicolor virginica
0.928 0.072
why the voting is not the same for the same data? what do i do wrong?
It's because the $votes components reflects the OOB predictions, whereas predict() gives you predictions based on all of the trees in the forest.
another 2 questions:
1. i tries to debug another problem in which the individual vector was
smaller the tree number in the forest.
i noticed that in this row of code:
treepred <- matrix(object$classes[t1$treepred], nrow =
length(keep),
dimnames = list(rn[keep], NULL))
the t1$treepred has values of 0 (i have 2 classes) and they
droped from the
results
what does this 0 mean?
Not sure why you're debugging that portion of the code. That is just to dimension the array passed back from C into a matrix. What is "t1"?
2. how can i drop a tree from the forest?
Look at the $forest component of the randomForest object, and subset the dimension that correspond to ntree in all of its components. Change $ntree accordingly. Andy
Thanks, Dror -- View this message in context: http://n4.nabble.com/Random-Forest-prediction-questions-tp1573 530p1573530.html Sent from the R help mailing list archive at Nabble.com.
______________________________________________ R-help at r-project.org mailing list 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.
Notice: This e-mail message, together with any attachme...{{dropped:10}}