lm_cv5.R

gistfile1.txt

data = read.csv("mydata.csv")

n = nrow(data) 
nfolds = 5 # 5-fold CV
groups = rep(1:nfolds, length=n)
set.seed(2)
cvgroups = sample(groups, n)

allpredicted = rep(NA, n)

for (ii in 1:nfolds) {
    group_ii = (cvgroups == ii) 
    train_data = data[!group_ii,]
    test_data = data[group_ii,]    
    model_fit = lm(model, data = train_data)
    predicted = predict(model_fit, newdata = test_data)
    allpredicted[group_ii] = predicted
}

# Compute CV value
y = data$y 
cv_value = mean((y - allpredicted)^2)

# Print CV value
print(cv_value)