# TODO: 导入必要的库和模块  
import numpy as np  
import pickle  
import matplotlib.pyplot as plt  
from sklearn.datasets import load_digits  
from sklearn.model_selection import train_test_split  
from sklearn.neighbors import KNeighborsClassifier  
from sklearn.metrics import accuracy_score  
from tqdm import tqdm  # 导入进度条库  

# TODO: 加载数字数据集  
digits = load_digits()  
X, y = digits.data, digits.target  

# TODO: 将数据集划分为训练集和测试集  
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)  

# TODO: 初始化变量以存储最佳准确率，相应的k值和最佳knn模型  
best_accuracy = 0  
best_k = 1  
best_model = None  

# TODO: 初始化一个列表以存储每个k值的准确率  
accuracies = []  

# TODO: 尝试从1到40的k值，对于每个k值，训练knn模型，保存最佳准确率，k值和knn模型  
for k in tqdm(range(1, 41,2), desc="Training KNN Models"):  
    # 创建KNN分类器  
    knn = KNeighborsClassifier(n_neighbors=k)  
    
    # 训练模型  
    knn.fit(X_train, y_train)  
    
    # 预测测试集  
    y_pred = knn.predict(X_test)  
    
    # 计算准确率  
    accuracy = accuracy_score(y_test, y_pred)  
    accuracies.append(accuracy)  
    
    # 更新最佳K值和模型  
    if accuracy > best_accuracy:  
        best_accuracy = accuracy  
        best_k = k  
        best_model = knn  

# TODO: 将最佳KNN模型保存到二进制文件  
with open('best_knn_model.pkl', 'wb') as f:  
    pickle.dump(best_model, f)  

# TODO: 打印最佳准确率和相应的k值  
print(f"\nBest K: {best_k} with Accuracy: {best_accuracy:.4f}")  

# 绘制准确率变化图  
plt.figure(figsize=(10, 6))  
plt.plot(range(1, 41,2), accuracies, marker='o', linestyle='-', color='b', label='Accuracy')  
plt.axvline(x=best_k, color='r', linestyle='--', label='Best K')  
plt.scatter(best_k, best_accuracy, color='red')  

# 添加文本标注  
plt.text(best_k, best_accuracy, f'K={best_k}\nAccuracy={best_accuracy:.2f}',   
         horizontalalignment='left', verticalalignment='bottom', fontsize=12, color='red')  

# 设置图表标题和标签  
plt.title('Accuracy of different k values', fontsize=16)  
plt.xlabel('k value', fontsize=14)  
plt.ylabel('Accuracy', fontsize=14)  
plt.xticks(range(1, 41))  
plt.ylim(0.965, 1)  # 根据示例调整y轴范围  
plt.grid()  
plt.legend()  

# 保存图形为PDF文件  
plt.savefig('accuracy_plot.pdf')  
plt.close()  

print("Accuracy plot saved to 'accuracy_plot.pdf'")