请用测试集训练集的写法，改写以下代码：library(caret) ctrl <- trainControl(method="cv",number=5) model1 <- train(data,y1,method="lm",trControl=ctrl) model2 <- train(data,y2,method="lm",trControl=ctrl) model3 <- train(data,y3,method="lm",trControl=ctrl) cv1 <- model1$results$RMSE cv2 <- model2$results$RMSE cv3 <- model3$results$RMSE set.seed(123) library(glmnet) x <- model.matrix(~.,data) y <- y1 # y2 或 y3 ridge1 <- cv.glmnet(x,y,alpha=0) ridge2 <- cv.glmnet(x,y,alpha=0) ridge3 <- cv.glmnet(x,y,alpha=0) plot(ridge1) plot(ridge2) plot(ridge3) set.seed(123) upper_bound <- mean(y1) + sd(y1)

以下是使用测试集训练集的代码：

set.seed(123)
library(caret)
trainIndex <- createDataPartition(y1, p=0.8, list=FALSE)
data_train <- data[trainIndex,]
data_test <- data[-trainIndex,]
y1_train <- y1[trainIndex]
y1_test <- y1[-trainIndex]
y2_train <- y2[trainIndex]
y2_test <- y2[-trainIndex]
y3_train <- y3[trainIndex]
y3_test <- y3[-trainIndex]

ctrl <- trainControl(method="cv", number=5)
model1 <- train(data_train, y1_train, method="lm", trControl=ctrl)
model2 <- train(data_train, y2_train, method="lm", trControl=ctrl)
model3 <- train(data_train, y3_train, method="lm", trControl=ctrl)
cv1 <- sqrt(mean((predict(model1, data_test) - y1_test)^2))
cv2 <- sqrt(mean((predict(model2, data_test) - y2_test)^2))
cv3 <- sqrt(mean((predict(model3, data_test) - y3_test)^2))

set.seed(123)
library(glmnet)
x <- model.matrix(~., data_train)
y <- y1_train
ridge1 <- cv.glmnet(x, y, alpha=0, nfolds=5)
y <- y2_train
ridge2 <- cv.glmnet(x, y, alpha=0, nfolds=5)
y <- y3_train
ridge3 <- cv.glmnet(x, y, alpha=0, nfolds=5)
plot(ridge1)
plot(ridge2)
plot(ridge3)

set.seed(123)
upper_bound <- mean(y1_train) + sd(y1_train)