library(rpart)

library(rpart.plot)

data(iris)

set.seed(123)

# Using sample_frac directly to get a fraction of the data

training_indices <- sample(nrow(iris), 0.7 * nrow(iris))

train_data <- iris[training_indices, ]

test_data <- iris[-training_indices, ]

# Check the structure of train_data

#str(train_data)

treemodel <- rpart(Species ~., data = train_data, method = "class")

rpart.plot(treemodel)

prediction <- predict(treemodel, newdata = test_data)

print(prediction)

# Extract predicted class probabilities

predicted_class <- colnames(prediction)[max.col(prediction)]

print(predicted_class)

conf_matrix <- table(predicted_class, test_data$Species)

print(conf_matrix)

accuracy <- sum(diag(conf_matrix)) / sum(conf_matrix)

print(accuracy)