Decision Tree CART

CART = Classification and Regression Tree, and for classification problem, criterion="gini" uses the Gini impurity criterion.

CART = Classification and Regression Tree, and for regression problem, criterion="squared_error" uses the squared_error impurity criterion.

Import required libraries

In [1]:
import pandas as pd
from sklearn.tree import DecisionTreeClassifier
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

import matplotlib.pyplot as plt
from sklearn.tree import plot_tree

Load Input

In [2]:
# Read weather.csv
df = pd.read_csv("weather.csv")

print(df.head())
print(df.columns)

    Outlook Temperature Humidity  Windy Play
0     Sunny         Hot     High  False   No
1     Sunny         Hot     High   True   No
2  Overcast         Hot     High  False  Yes
3      Rain        Mild     High  False  Yes
4      Rain        Cool   Normal  False  Yes
Index(['Outlook', 'Temperature', 'Humidity', 'Windy', 'Play'], dtype='object')

# Encode categorical columns

In [3]:
# Create encoders
encoders = {}

# Encode categorical columns
for column in df.columns:
    le = LabelEncoder()
    df[column] = le.fit_transform(df[column])
    encoders[column] = le

# Separate input and output

In [4]:
X_features = df.drop("Play", axis=1)
y_target = df["Play"]

# Split the dataset

In [5]:
X_train, X_test, y_train, y_test = train_test_split(
    X_features,
    y_target,
    test_size=0.2,
    random_state=42
)

# Build CART Decision Tree

  • criterion="gini" → uses Gini impurity, the standard CART criterion for classification.
  • fit() → builds the decision tree.
  • predict() → predicts the class for new data.
In [6]:
model = DecisionTreeClassifier(
    criterion="gini",
    random_state=42
)

# Train the model

In [7]:
model.fit(X_train, y_train)
Out[7]:
DecisionTreeClassifier(random_state=42)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Prediction

In [8]:
y_pred = model.predict(X_test)

Accuracy

In [9]:
print("Predicted:", y_pred)
print("Actual:", y_test.values)
print("Accuracy:", accuracy_score(y_test, y_pred))

Predicted: [1 1 0]
Actual: [1 1 0]
Accuracy: 1.0

Visualize the CART tree

In [10]:
plt.figure(figsize=(18, 15))

plot_tree(
    model,
    feature_names=X_features.columns,
    class_names=encoders["Play"].classes_,
    filled=True,
    fontsize=18
)

plt.title("CART Decision Tree")
plt.show()
In [11]:
from sklearn import tree
clf=DecisionTreeClassifier(
    criterion="gini",
    random_state=42
)
model=clf.fit(X_train, y_train)
In [12]:
text_representation = tree.export_text(
    clf,
    feature_names=list(X_train.columns)
)
print(text_representation)

|--- Outlook <= 0.50
|   |--- class: 1
|--- Outlook >  0.50
|   |--- Humidity <= 0.50
|   |   |--- Windy <= 0.50
|   |   |   |--- Outlook <= 1.50
|   |   |   |   |--- class: 1
|   |   |   |--- Outlook >  1.50
|   |   |   |   |--- class: 0
|   |   |--- Windy >  0.50
|   |   |   |--- class: 0
|   |--- Humidity >  0.50
|   |   |--- Windy <= 0.50
|   |   |   |--- class: 1
|   |   |--- Windy >  0.50
|   |   |   |--- Outlook <= 1.50
|   |   |   |   |--- class: 0
|   |   |   |--- Outlook >  1.50
|   |   |   |   |--- class: 1

Make prediction for new data

In [13]:
new_data = pd.DataFrame([{
    "Outlook": "Sunny",
    "Temperature": "Cool",
    "Humidity": "High",
    "Windy": False
}])

Encode the new data

# Convert False/True to the same format as training data

In [14]:
new_data["Windy"] = new_data["Windy"].astype(str)

# Encode new data using the SAME encoders

In [15]:
# Encode new data using existing encoders
for column in X_features.columns:
    new_data[column] = encoders[column].transform(new_data[column])
In [16]:
# Predict
prediction = model.predict(new_data)
In [17]:
# Convert encoded prediction back to Yes/No
prediction_label = encoders["Play"].inverse_transform(prediction)
print("\nNew Data Prediction:", prediction_label[0])

New Data Prediction: No
In [ ]: