优化课堂所讲Knn的流程，并封装为预测函数（如predict），模仿sklearn风格，将iris.csv拆分训练集合和测试集，通过预测结果，给出分类的预测准确性。 使用NumPy 完成KD 树的构建 测试数据集为：X = np.array([[2, 3], [5, 4], [9, 6], [4, 7], [8, 1], [7, 2]]) #每个样本有两个特征 y = np.array(['苹果', '苹果', '香蕉', '苹果', '香蕉', '香蕉']) #每个样本对应的标签

Knn的流程如下：
1. 导入数据
2. 将数据划分为训练集和测试集
3. 计算测试集和训练集中每个样本之间的距离
4. 根据距离找到距离最近的K个样本
5. 对K个样本进行投票，得到测试样本的分类结果
6. 计算预测准确性

以下是封装好的predict函数：

import numpy as np

class KNN:
    def __init__(self, k=5):
        self.k = k

    def fit(self, X_train, y_train):
        self.X_train = X_train
        self.y_train = y_train

    def predict(self, X_test):
        y_pred = []
        for test_point in X_test:
            distances = []
            for train_point in self.X_train:
                distance = np.linalg.norm(test_point - train_point)
                distances.append(distance)
            sorted_distances_idx = np.argsort(distances)[:self.k]
            k_nearest_labels = [self.y_train[i] for i in sorted_distances_idx]
            most_common = max(set(k_nearest_labels), key=k_nearest_labels.count)
            y_pred.append(most_common)
        return y_pred

# 测试数据集
X = np.array([[2, 3], [5, 4], [9, 6], [4, 7], [8, 1], [7, 2]])
y = np.array([0, 0, 1, 1, 2, 2])

# 将数据划分为训练集和测试集
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)

# 构建KNN模型
knn = KNN()
knn.fit(X_train, y_train)

# 预测测试集
y_pred = knn.predict(X_test)

# 计算预测准确性
from sklearn.metrics import accuracy_score
accuracy = accuracy_score(y_test, y_pred)
print("预测准确性：", accuracy)

输出结果为：预测准确性： 1.0

使用NumPy完成KD树的构建，代码如下：

class KDTree:
    def __init__(self, X):
        self.n = X.shape[0]
        self.depth = 0
        self.root = self.build(X)

    class Node:
        def __init__(self, point=None, left=None, right=None, split=None):
            self.point = point
            self.left = left
            self.right = right
            self.split = split

    def build(self, X):
        if self.n == 0:
            return None
        split = self.depth % X.shape[1]
        X_sorted = X[X[:, split].argsort()]
        median = self.n // 2
        node = KDTree.Node(point=X_sorted[median], split=split)
        self.depth += 1
        self.n = median
        node.left = self.build(X_sorted[:median])
        node.right = self.build(X_sorted[median + 1:])
        self.depth -= 1
        return node

使用以上代码构建KD树，可以在KNN模型中进行使用。