tree-cricket-depth.py

import sklearn.tree
import sklearn.metrics
import matplotlib.pyplot as plt
import sklearn.tree
import sklearn.preprocessing
import pandas as pd

df = pd.read_csv('./data/cricket.csv')
labelEncoder = sklearn.preprocessing.LabelEncoder()
features = ['outlook', 'temperature', 'humidity', 'wind']

for col in df[features]:
    df[col] = labelEncoder.fit_transform(df[col])
decisionTreeClassifier = sklearn.tree.DecisionTreeClassifier(
    criterion    = 'entropy',
    random_state = 42,
)

# since small data, no split
decisionTreeClassifier.fit(X=df[features], y=df['play'])
fig, axes = plt.subplots(2, 2, layout='tight')
for i in range(4):
    decisionTreeClassifier = sklearn.tree.DecisionTreeClassifier(
        criterion    = 'entropy',
        random_state = 42,
        max_depth    = i + 1,
    )
    decisionTreeClassifier.fit(X=df[features], y=df['play'])
    predict = decisionTreeClassifier.predict(df[features])
    accuracy = sklearn.metrics.accuracy_score(df['play'], predict)

    ax = axes.ravel()[i]
    ax.set_title('max_depth={} accuracy={}'.format(i+1, accuracy))
    sklearn.tree.plot_tree(
        decisionTreeClassifier,
        ax            = ax,
        feature_names = features,
        filled        = True,
        class_names   = None,  # ['✗', '✓'],  # handled by filled color
        rounded       = True,
        fontsize      = 8,
        impurity      = False,
    )
plt.show()