r语言建立bp错误提示 REAL() can only be applied to a 'numeric', not a

麻烦各位大神帮帮忙~在用r软件建立bp神经网络时遇到问题~
r语言源代码为：
建立bp网络：
net <- newff(n.neurons=c(50,8,2,1), learning.rate.global=1e-2, momentum.global=0.5,
        error.criterium="LMS", Stao=NA, hidden.layer="tansig",
        output.layer="purelin", method="ADAPTgdwm")
（1）训练bp网络
result <- train(net, trainset[,-51],trainset$y, error.criterium="LMS", report=TRUE, show.step=100, n.shows=5 )
错误提示：
Error in train.method(net, P, T, show.step, n.threads = n.threads) :
  REAL() can only be applied to a 'numeric', not a 'character'
（2）采用x=matrix(as.numeric(x),nrow=nrow(x))语句想将字符型改为数字型出现错误提示：
Error in matrix(as.numeric(x), nrow = nrow(x)) :   (list) object cannot be coerced to type 'double'
（3）模型之前想对数据进行标准化用语句
scale(x, scale = TRUE)出现错误提示：Error in colMeans(x, na.rm = TRUE) : 'x' must be numeric。

原来是x的格式是list，要先转换格式后在标准化
x=matrix(unlist(x),nrow=nrow(x))
x=scale(x, scale = TRUE)
index = 1:nrow(x)#行数
testindex = sample(index, trunc(length(index)/3))#抽样的过程，抽取测试集的行号，从index中抽取约1/3的样本，trunc()向零取整
testset = x[testindex, ]#testset测试集
trainset = x[-testindex, ]#trainset训练集
net <- newff(n.neurons=c(50,8,2,1), learning.rate.global=1e-2, momentum.global=0.5,
        error.criterium="LMS", Stao=NA, hidden.layer="tansig",
        output.layer="purelin", method="ADAPTgdwm")

result <- train(net, trainset[,-51],trainset[,51], error.criterium="LMS", report=TRUE, show.step=100, n.shows=5 )

net <- newff(n.neurons=c(50,8,2,1), learning.rate.global=1e-2, momentum.global=0.5,
        error.criterium="LMS", Stao=NA, hidden.layer="tansig",
        output.layer="purelin", method="ADAPTgdwm")

result <- train(net, trainset[,-51],trainset[,51], error.criterium="LMS", report=TRUE, show.step=100, n.shows=5 )
后面这是什么意思？