Decision Trees

import numpy as np
import pandas as pd
from sklearn.tree import DecisionTreeClassifier, export_text
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.metrics import accuracy_score, classification_report
from sklearn.datasets import load_breast_cancer

print("=" * 55)
print("Decision Trees: كشف سرطان الثدي")
print("=" * 55)

# بيانات طبية حقيقية (مدمجة في sklearn)
data = load_breast_cancer()
X = pd.DataFrame(data.data, columns=data.feature_names)
y = data.target

print(f"عدد العينات: {len(y)}")
print(f"الفئات: {dict(zip(data.target_names, np.bincount(y)))}")

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, stratify=y, random_state=42
)

# ─────────────────────────────────────────
# 1. شجرة بدون قيود — Overfitting
# ─────────────────────────────────────────
tree_full = DecisionTreeClassifier(random_state=42)
tree_full.fit(X_train, y_train)

train_acc = accuracy_score(y_train, tree_full.predict(X_train))
test_acc  = accuracy_score(y_test,  tree_full.predict(X_test))

print("\n1. شجرة كاملة (بدون قيود):")
print(f"   Training Accuracy = {train_acc:.3f}")
print(f"   Test Accuracy     = {test_acc:.3f}")
print(f"   عمق الشجرة        = {tree_full.get_depth()}")
print(f"   Overfitting gap   = {train_acc - test_acc:.3f}")

# ─────────────────────────────────────────
# 2. شجرة مقلّمة — Pruning
# ─────────────────────────────────────────
tree_pruned = DecisionTreeClassifier(
    max_depth=5,
    min_samples_leaf=10,
    min_samples_split=20,
    random_state=42
)
tree_pruned.fit(X_train, y_train)

train_acc_p = accuracy_score(y_train, tree_pruned.predict(X_train))
test_acc_p  = accuracy_score(y_test,  tree_pruned.predict(X_test))

print("\n2. شجرة مقلّمة (max_depth=5):")
print(f"   Training Accuracy = {train_acc_p:.3f}")
print(f"   Test Accuracy     = {test_acc_p:.3f}")
print(f"   عمق الشجرة        = {tree_pruned.get_depth()}")
print(f"   Overfitting gap   = {train_acc_p - test_acc_p:.3f}  ✅ أفضل!")

# ─────────────────────────────────────────
# 3. Cross-Validation
# ─────────────────────────────────────────
cv_scores = cross_val_score(tree_pruned, X, y, cv=5, scoring="accuracy")
print(f"\n3. Cross-Validation (5 folds):")
print(f"   Scores: {cv_scores.round(3)}")
print(f"   Mean:   {cv_scores.mean():.3f} ± {cv_scores.std():.3f}")

# ─────────────────────────────────────────
# 4. أهمية الميزات
# ─────────────────────────────────────────
importances = pd.Series(
    tree_pruned.feature_importances_,
    index=data.feature_names
).sort_values(ascending=False)

print("\n4. أهم الميزات الطبية:")
for feat, imp in importances.head(5).items():
    bar = "█" * int(imp * 50)
    print(f"   {feat[:30]:30s}: {imp:.3f} {bar}")

# ─────────────────────────────────────────
# 5. قراءة الشجرة
# ─────────────────────────────────────────
print("\n5. هيكل الشجرة (أول 3 مستويات):")
tree_text = export_text(tree_pruned,
                        feature_names=list(data.feature_names),
                        max_depth=3)
print(tree_text[:800])