※ 引述《Psycap (siraPcitsatnaF)》之銘言：
: ※ 引述《litchi (狂戀大屁股)》之銘言：
: Do you run this?
: model <- randomForest(traindata[,-1],factor(traindata[,1]))
No. Because I just want to show the difference of memory usage
between data.frame and matrix.
: The size of traindata was about 56mb 38mb, but as the above was executing,
: it's memory used went up to 2234m 934m.
: Why?
No only the data set but also randomForest needs much memory. Building
500 trees for the huge ijcnn1 data set should take pretty much memory.
I think the situation is quite reasonable. :)