#read in data and set factors/numerics pen = read.table('pen2ref.txt',header=T) facind = c( 1, 3, 5, 7, 9, 19, 20, 21) #which variables are factors (categorical) for(i in facind) pen[[i]] = as.factor(pen[[i]]) #train,val,test set.seed(99) n=nrow(pen) n1=floor(n/2) n2=floor(n/4) n3=n-n1-n2 ii = sample(1:n,n) pentrain=pen[ii[1:n1],] penval = pen[ii[n1+1:n2],] pentest = pen[ii[n1+n2+1:n3],]