How to solve "cannot coerce class to data.frame? - r

Question

Problems occur in Line 20: x3 <- lm(Salary ~ ...
Error in as.data.frame.default(data) : cannot coerce class ‘c("train", "train.formula")’ to a data.frame
How to solve?
attach(Hitters)
Hitters
library(caret)
set.seed(123)
# Define training control
set.seed(123)
train.control <- trainControl(method = "cv", number = 10)
# Train the model
x2 <- train(Salary ~., data = x, method = "lm",
trControl = train.control)
# Summarize the results
print(x)
x3 <- lm(Salary ~ poly(AtBat,3) + poly(Hits,3) + poly(Walks,3) + poly(CRuns,3) + poly(CWalks,3) + poly(PutOuts,3), data = x2)
summary(x3)
MSE = mean(x3$residuals^2)
print("Mean Squared Error: ")
print(MSE)

Answer 1

First, as #dcarlson already mentioned, you should define x.
Second, x3 does not return a data frame.
If you run
str(x2)
you'll see that all the elements you're using in the lm function are part of a data frame called trainingData.
So if you intend to use the lm function, use that as your data source in the lm function, NOT x2.
I've rewritten your code below.
PS I'm far from a R expert so if someone wants to shoot at this answer, go ahead, I'm always willing to learn ;)
attach(Hitters)
Hitters
library(caret)
set.seed(123)
# Define training control
set.seed(123)
train.control <- trainControl(method = "cv", number = 10)
# Train the model
x2 <- train(Salary ~., data = x, method = "lm", trControl = train.control)
# Summarize the results
print(x2)
# str(x2) # $trainingData data.frame
x2$trainingData[["AtBat"]]
m <- x2$trainingData
x3 <- lm(Salary ~ poly(AtBat,3) + poly(Hits,3) + poly(Walks,3) + poly(CRuns,3) + poly(CWalks,3) + poly(PutOuts,3), data = m)
summary(x3)
MSE = mean(x3$residuals^2)
cat("Mean Squared Error: ", MSE) # use cat to concatenate text and variable value in one line