library(ElemStatLearn)
data(spam)
trans<-apply(spam[,1:57],2,function (x) { log(x+0.1) })
norma <- apply(trans, 2, function (x) { (x-mean(x))/sd(x) })
spam_trans<-cbind(norma,spam[,58])
training <- sample(1:(dim(spam)[1]), 2301)
actuals <- spam[-training,58]
qda_fit<-qda(norma[training,],spam$spam[training])

提醒 ：“  错误于qda.default(x, grouping, ...) : rank deficiency in group spam”  
请问高手们是什么原因啊？怎样才能得到正确的   QDA Analysis    ？？？？