How to Use Decision Tree in R: Simple Guide with Example

To use a decision tree in R, install and load the rpart package, then create a model using rpart() with a formula and dataset. You can visualize the tree with rpart.plot or plot() functions.

Syntax

The basic syntax to create a decision tree model in R using the rpart package is:

• rpart(formula, data, method)

Where:

• formula defines the target and predictors (e.g., target ~ feature1 + feature2).
• data is the dataset used for training.
• method specifies the type of problem: "class" for classification or "anova" for regression.

r

library(rpart)
model <- rpart(target ~ feature1 + feature2, data = dataset, method = "class")

Example

This example shows how to build a decision tree to classify the species in the famous Iris dataset. It trains the model and plots the tree.

r

library(rpart)
library(rpart.plot)

# Load iris dataset
data(iris)

# Build decision tree model to predict Species
model <- rpart(Species ~ ., data = iris, method = "class")

# Print model summary
print(model)

# Plot the decision tree
rpart.plot(model)

Output

n= 150 

node), split, n, loss, yval, (yprob)
      * denotes terminal node

1) root 150 100 setosa (0.33333333 0.33333333 0.33333333)  
  2) Petal.Length< 2.45 50 0 setosa (1.00000000 0.00000000 0.00000000) *
  3) Petal.Length>=2.45 100 50 versicolor (0.00000000 0.50000000 0.50000000)  
    6) Petal.Width< 1.75 54 5 versicolor (0.00000000 0.90740741 0.09259259)  
     12) Petal.Length< 4.95 48 1 versicolor (0.00000000 0.97916667 0.02083333) *
     13) Petal.Length>=4.95 6 0 virginica (0.00000000 0.00000000 1.00000000) *
    7) Petal.Width>=1.75 46 1 virginica (0.00000000 0.02173913 0.97826087) *

Common Pitfalls

Common mistakes when using decision trees in R include:

• Not loading the rpart package before using rpart().
• Using incorrect formula syntax or missing the target variable.
• Forgetting to specify method = "class" for classification problems.
• Not installing or loading rpart.plot for visualization.

Always check your data and formula carefully.

r

## Wrong: missing method for classification
model_wrong <- rpart(Species ~ ., data = iris)

## Right: specify method
model_right <- rpart(Species ~ ., data = iris, method = "class")

Quick Reference

Function	Purpose	Notes
rpart()	Create decision tree model	Use method="class" for classification
rpart.plot()	Plot decision tree	Requires rpart.plot package
print()	Show model summary	Shows splits and node info
predict()	Make predictions	Use type="class" for class labels

Key Takeaways

Use the rpart package and rpart() function to build decision trees in R.

Specify method="class" for classification problems to get correct results.

Visualize trees easily with rpart.plot() from the rpart.plot package.

Check your formula and data carefully to avoid common errors.

Use predict() with type="class" to get predicted classes from the model.