Random Forest

import numpy as np
import pandas as pd
from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.metrics import accuracy_score, classification_report
from sklearn.datasets import make_classification

print("=" * 55)
print("Random Forest: المقارنة مع Decision Tree")
print("=" * 55)

# بيانات معقدة
X, y = make_classification(
    n_samples=2000, n_features=20, n_informative=10,
    n_redundant=5, random_state=42
)

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

# ─────────────────────────────────────────
# 1. Decision Tree للمقارنة
# ─────────────────────────────────────────
dt = DecisionTreeClassifier(random_state=42)
dt.fit(X_train, y_train)

dt_train = accuracy_score(y_train, dt.predict(X_train))
dt_test  = accuracy_score(y_test,  dt.predict(X_test))

print(f"\nDecision Tree:")
print(f"  Train Accuracy: {dt_train:.3f}")
print(f"  Test  Accuracy: {dt_test:.3f}")
print(f"  Overfit gap:    {dt_train - dt_test:.3f}")

# ─────────────────────────────────────────
# 2. Random Forest
# ─────────────────────────────────────────
rf = RandomForestClassifier(
    n_estimators=200,
    max_features="sqrt",
    oob_score=True,
    n_jobs=-1,
    random_state=42
)
rf.fit(X_train, y_train)

rf_train = accuracy_score(y_train, rf.predict(X_train))
rf_test  = accuracy_score(y_test,  rf.predict(X_test))

print(f"\nRandom Forest (200 trees):")
print(f"  Train Accuracy: {rf_train:.3f}")
print(f"  Test  Accuracy: {rf_test:.3f}")
print(f"  OOB Score:      {rf.oob_score_:.3f}  ← مجاني!")
print(f"  Overfit gap:    {rf_train - rf_test:.3f}")

# ─────────────────────────────────────────
# 3. تأثير عدد الأشجار
# ─────────────────────────────────────────
print("\n3. تأثير عدد الأشجار على الدقة:")
for n in [1, 5, 10, 50, 100, 200, 500]:
    rf_n = RandomForestClassifier(n_estimators=n, n_jobs=-1, random_state=42)
    rf_n.fit(X_train, y_train)
    acc = accuracy_score(y_test, rf_n.predict(X_test))
    bar = "█" * int(acc * 30)
    print(f"  n={n:4d}: {acc:.3f} {bar}")

# ─────────────────────────────────────────
# 4. أهمية الميزات
# ─────────────────────────────────────────
importance_df = pd.DataFrame({
    "feature": [f"feature_{i}" for i in range(20)],
    "importance": rf.feature_importances_
}).sort_values("importance", ascending=False)

print("\n4. أهم 10 ميزات:")
for _, row in importance_df.head(10).iterrows():
    bar = "█" * int(row["importance"] * 200)
    print(f"  {row['feature']:12s}: {row['importance']:.4f} {bar}")