# 示例：展示集成學習的優勢
import numpy as np
import matplotlib.pyplot as plt
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_moons, make_circles
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
# 設置中文字體
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False

def improved_ensemble_demo():
    
    # 創建更復雜的非線性數據集
    print(" 生成複雜數據集...")
    X, y = make_moons(n_samples=300, noise=0.3, random_state=42)
    
    # 分割數據
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
    
    # 單個深度決策樹（容易過擬合）
    single_tree = DecisionTreeClassifier(max_depth=15, random_state=42)
    # 隨機森林（集成方法）
    random_forest = RandomForestClassifier(n_estimators=50, max_depth=5, random_state=42)
    
    # 訓練模型
    print(" 訓練模型...")
    single_tree.fit(X_train, y_train)
    random_forest.fit(X_train, y_train)
    
    # 計算準確率
    single_train_acc = accuracy_score(y_train, single_tree.predict(X_train))
    single_test_acc = accuracy_score(y_test, single_tree.predict(X_test))
    rf_train_acc = accuracy_score(y_train, random_forest.predict(X_train))
    rf_test_acc = accuracy_score(y_test, random_forest.predict(X_test))
    
    print(f"\n 性能對比:")
    print(f"單個決策樹 - 訓練準確率: {single_train_acc:.3f}, 測試準確率: {single_test_acc:.3f}")
    print(f"隨機森林   - 訓練準確率: {rf_train_acc:.3f}, 測試準確率: {rf_test_acc:.3f}")
    
    # 創建更密集的網格來可視化決策邊界
    x_min, x_max = X[:, 0].min() - 0.5, X[:, 0].max() + 0.5
    y_min, y_max = X[:, 1].min() - 0.5, X[:, 1].max() + 0.5
    xx, yy = np.meshgrid(np.arange(x_min, x_max, 0.02),
                         np.arange(y_min, y_max, 0.02))
    
    # 創建圖形
    plt.figure(figsize=(15, 5))
    
    # 子圖1：原始數據
    plt.subplot(1, 3, 1)
    scatter = plt.scatter(X[:, 0], X[:, 1], c=y, s=30, cmap=plt.cm.RdYlBu, edgecolors='k')
    plt.xlabel('特徵 1')
    plt.ylabel('特徵 2')
    plt.title('原始數據分佈\n(月亮數據集)')
    plt.colorbar(scatter)
    plt.grid(True, alpha=0.3)
    
    # 子圖2：單個決策樹的決策邊界
    plt.subplot(1, 3, 2)
    Z_single = single_tree.predict(np.c_[xx.ravel(), yy.ravel()])
    Z_single = Z_single.reshape(xx.shape)
    plt.contourf(xx, yy, Z_single, alpha=0.8, cmap=plt.cm.RdYlBu)
    plt.scatter(X[:, 0], X[:, 1], c=y, s=30, cmap=plt.cm.RdYlBu, edgecolors='k')
    plt.xlabel('特徵 1')
    plt.ylabel('特徵 2')
    plt.title(f'單個決策樹 (深度=15)\n訓練準確率: {single_train_acc:.3f}, 測試準確率: {single_test_acc:.3f}')
    plt.grid(True, alpha=0.3)
    
    # 子圖3：隨機森林的決策邊界
    plt.subplot(1, 3, 3)
    Z_ensemble = random_forest.predict(np.c_[xx.ravel(), yy.ravel()])
    Z_ensemble = Z_ensemble.reshape(xx.shape)
    plt.contourf(xx, yy, Z_ensemble, alpha=0.8, cmap=plt.cm.RdYlBu)
    plt.scatter(X[:, 0], X[:, 1], c=y, s=30, cmap=plt.cm.RdYlBu, edgecolors='k')
    plt.xlabel('特徵 1')
    plt.ylabel('特徵 2')
    plt.title(f'隨機森林 (50棵樹)\n訓練準確率: {rf_train_acc:.3f}, 測試準確率: {rf_test_acc:.3f}')
    plt.grid(True, alpha=0.3)
    
    plt.tight_layout()
    plt.show()
    
    # 添加過擬合分析
    print(f"\n 過擬合分析:")
    overfitting_single = single_train_acc - single_test_acc
    overfitting_rf = rf_train_acc - rf_test_acc
    print(f"單個決策樹過擬合程度: {overfitting_single:.3f}")
    print(f"隨機森林過擬合程度: {overfitting_rf:.3f}")
    
    if overfitting_single > overfitting_rf:
        print(" 隨機森林有效減少了過擬合！")
    else:
        print(" 在這個例子中過擬合改善不明顯")

improved_ensemble_demo()

# 示例：圓形數據集展示集成學習的優勢
import numpy as np
import matplotlib.pyplot as plt
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_moons, make_circles
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

# 設置中文字體
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False


def circles_ensemble_demo():
    """使用圓形數據集展示集成學習的優勢"""
    
    # 創建圓形數據集（更難分類）
    print(" 生成圓形數據集...")
    X, y = make_circles(n_samples=400, noise=0.2, factor=0.5, random_state=42)
    
    # 分割數據
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
    
    # 創建不同複雜度的模型
    models = {
        '簡單決策樹': DecisionTreeClassifier(max_depth=3, random_state=42),
        '複雜決策樹': DecisionTreeClassifier(max_depth=20, random_state=42),
        '隨機森林': RandomForestClassifier(n_estimators=100, max_depth=5, random_state=42)
    }
    
    # 訓練並評估所有模型
    results = {}
    
    print(" 訓練和評估模型...")
    for name, model in models.items():
        model.fit(X_train, y_train)
        train_acc = accuracy_score(y_train, model.predict(X_train))
        test_acc = accuracy_score(y_test, model.predict(X_test))
        results[name] = {'model': model, 'train_acc': train_acc, 'test_acc': test_acc}
        print(f"{name:12} - 訓練: {train_acc:.3f}, 測試: {test_acc:.3f}")
    
    # 創建可視化網格
    x_min, x_max = X[:, 0].min() - 0.5, X[:, 0].max() + 0.5
    y_min, y_max = X[:, 1].min() - 0.5, X[:, 1].max() + 0.5
    xx, yy = np.meshgrid(np.arange(x_min, x_max, 0.02),
                         np.arange(y_min, y_max, 0.02))
    
    # 創建對比圖
    plt.figure(figsize=(18, 5))
    
    # 原始數據
    plt.subplot(1, 4, 1)
    scatter = plt.scatter(X[:, 0], X[:, 1], c=y, s=30, cmap=plt.cm.RdYlBu, edgecolors='k')
    plt.xlabel('特徵 1')
    plt.ylabel('特徵 2')
    plt.title('原始數據分佈\n(圓形數據集)')
    plt.colorbar(scatter)
    plt.grid(True, alpha=0.3)
    
    # 三個模型的決策邊界
    for i, (name, result) in enumerate(results.items()):
        plt.subplot(1, 4, i+2)
        Z = result['model'].predict(np.c_[xx.ravel(), yy.ravel()])
        Z = Z.reshape(xx.shape)
        plt.contourf(xx, yy, Z, alpha=0.8, cmap=plt.cm.RdYlBu)
        plt.scatter(X[:, 0], X[:, 1], c=y, s=30, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.6)
        plt.xlabel('特徵 1')
        plt.ylabel('特徵 2')
        
        overfitting = result['train_acc'] - result['test_acc']
        plt.title(f'{name}\n訓練: {result["train_acc"]:.3f}, 測試: {result["test_acc"]:.3f}\n過擬合: {overfitting:.3f}')
        plt.grid(True, alpha=0.3)
    
    plt.tight_layout()
    plt.show()
    
    # 性能總結
    print(f"\n 性能總結:")
    best_model = max(results.items(), key=lambda x: x[1]['test_acc'])
    print(f"最佳模型: {best_model[0]} (測試準確率: {best_model[1]['test_acc']:.3f})")

# 運行圓形數據集示例
circles_ensemble_demo()

# 示例：投票機制演示展示集成學習的優勢
import numpy as np
import matplotlib.pyplot as plt
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_moons, make_circles
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

# 設置中文字體
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False

def voting_mechanism_demo_improved():
    """投票機制演示"""
    
    # 創建簡單的二分類問題
    np.random.seed(42)
    X = np.random.randn(200, 2)
    y = (X[:, 0] ** 2 + X[:, 1] ** 2 > 1.5).astype(int)
    
    # 創建5個不同的弱分類器
    from sklearn.linear_model import LogisticRegression
    from sklearn.svm import SVC
    from sklearn.neighbors import KNeighborsClassifier
    
    classifiers = [
        ('決策樹1', DecisionTreeClassifier(max_depth=2, random_state=1)),
        ('決策樹2', DecisionTreeClassifier(max_depth=2, random_state=2)),
        ('決策樹3', DecisionTreeClassifier(max_depth=2, random_state=3)),
        ('邏輯迴歸', LogisticRegression(random_state=42)),
        ('K近鄰', KNeighborsClassifier(n_neighbors=3))
    ]
    
    # 訓練所有分類器
    individual_predictions = []
    individual_accuracies = []
    
    print(" 集成學習的投票機制演示")
    print("=" * 40)
    
    for name, clf in classifiers:
        clf.fit(X, y)
        y_pred = clf.predict(X)
        acc = accuracy_score(y, y_pred)
        individual_predictions.append(y_pred)
        individual_accuracies.append(acc)
        print(f"{name:10} - 準確率: {acc:.3f}")
    
    # 集成預測：多數投票
    individual_predictions = np.array(individual_predictions)
    ensemble_pred = (np.sum(individual_predictions, axis=0) > len(classifiers) / 2).astype(int)
    ensemble_acc = accuracy_score(y, ensemble_pred)
    
    print(f"{'集成投票':10} - 準確率: {ensemble_acc:.3f}")
    print(f" 集成相比平均提升: {ensemble_acc - np.mean(individual_accuracies):.3f}")
    
    # 創建上3下3的佈局
    plt.figure(figsize=(18, 12))
    
    # 第一行：原始數據 + 兩個分類器
    # 子圖1：原始數據
    plt.subplot(2, 3, 1)
    scatter = plt.scatter(X[:, 0], X[:, 1], c=y, s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title('原始數據分佈\n(真實標籤)', fontsize=14, fontweight='bold')
    plt.xlabel('特徵1', fontsize=12)
    plt.ylabel('特徵2', fontsize=12)
    plt.grid(True, alpha=0.3)
    
    # 子圖2：決策樹1
    plt.subplot(2, 3, 2)
    plt.scatter(X[:, 0], X[:, 1], c=individual_predictions[0], s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title(f'決策樹1預測\n準確率: {individual_accuracies[0]:.3f}', fontsize=14)
    plt.xlabel('特徵1', fontsize=12)
    plt.ylabel('特徵2', fontsize=12)
    plt.grid(True, alpha=0.3)
    
    # 子圖3：決策樹2
    plt.subplot(2, 3, 3)
    plt.scatter(X[:, 0], X[:, 1], c=individual_predictions[1], s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title(f'決策樹2預測\n準確率: {individual_accuracies[1]:.3f}', fontsize=14)
    plt.xlabel('特徵1', fontsize=12)
    plt.ylabel('特徵2', fontsize=12)
    plt.grid(True, alpha=0.3)
    
    # 第二行：另外兩個分類器 + 集成結果
    # 子圖4：決策樹3
    plt.subplot(2, 3, 4)
    plt.scatter(X[:, 0], X[:, 1], c=individual_predictions[2], s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title(f'決策樹3預測\n準確率: {individual_accuracies[2]:.3f}', fontsize=14)
    plt.xlabel('特徵1', fontsize=12)
    plt.ylabel('特徵2', fontsize=12)
    plt.grid(True, alpha=0.3)
    
    # 子圖5：邏輯迴歸
    plt.subplot(2, 3, 5)
    plt.scatter(X[:, 0], X[:, 1], c=individual_predictions[3], s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title(f'邏輯迴歸預測\n準確率: {individual_accuracies[3]:.3f}', fontsize=14)
    plt.xlabel('特徵1', fontsize=12)
    plt.ylabel('特徵2', fontsize=12)
    plt.grid(True, alpha=0.3)
    
    # 子圖6：K近鄰
    plt.subplot(2, 3, 6)
    plt.scatter(X[:, 0], X[:, 1], c=individual_predictions[4], s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title(f'K近鄰預測\n準確率: {individual_accuracies[4]:.3f}', fontsize=14)
    plt.xlabel('特徵1', fontsize=12)
    plt.ylabel('特徵2', fontsize=12)
    plt.grid(True, alpha=0.3)
    
    plt.tight_layout()
    plt.show()
    
    # 單獨顯示集成結果
    plt.figure(figsize=(8, 6))
    plt.scatter(X[:, 0], X[:, 1], c=ensemble_pred, s=50, cmap=plt.cm.RdYlBu, edgecolors='k', alpha=0.8)
    plt.title(f'集成投票結果\n準確率: {ensemble_acc:.3f} (相比平均提升: {ensemble_acc - np.mean(individual_accuracies):.3f})', 
              fontsize=16, fontweight='bold', pad=20)
    plt.xlabel('特徵1', fontsize=14)
    plt.ylabel('特徵2', fontsize=14)
    plt.grid(True, alpha=0.3)
    plt.tight_layout()
    plt.show()
    
    # 展示投票過程
    print(f"\n 投票過程示例 (前10個樣本):")
    print("樣本 |", " | ".join([name for name, _ in classifiers]), "| 集成結果 | 真實標籤 | 是否正確")
    print("-" * 85)
    
    correct_count = 0
    for i in range(10):
        votes = individual_predictions[:, i]
        vote_counts = np.sum(votes)
        result = 1 if vote_counts > len(classifiers) / 2 else 0
        is_correct = "" if result == y[i] else "❌"
        if result == y[i]:
            correct_count += 1
        
        print(f"{i:4} |", " | ".join([f"{vote:8}" for vote in votes]), 
              f"| {result:8} | {y[i]:8} | {is_correct}")
    
    print(f"\n 前10個樣本集成準確率: {correct_count}/10 = {correct_count/10:.1%}")
    
    # 添加性能分析
    print(f"\n 詳細性能分析:")
    print(f"單個模型平均準確率: {np.mean(individual_accuracies):.3f}")
    print(f"集成模型準確率: {ensemble_acc:.3f}")
    print(f"相對提升: {(ensemble_acc - np.mean(individual_accuracies)) / np.mean(individual_accuracies) * 100:.1f}%")
    
    # 計算多樣性指標
    diversity_scores = []
    for i in range(len(classifiers)):
        for j in range(i+1, len(classifiers)):
            # 計算兩個分類器預測不同的比例
            disagreement = np.mean(individual_predictions[i] != individual_predictions[j])
            diversity_scores.append(disagreement)
            print(f"{classifiers[i][0]} vs {classifiers[j][0]} 差異度: {disagreement:.3f}")
    
    print(f"平均模型差異度: {np.mean(diversity_scores):.3f}")
    print(" 關鍵洞察: 模型差異度越高，集成效果通常越好！")

# 運行改進的投票機制演示
voting_mechanism_demo_improved()