不调用工具包，实现Kmeans聚类，改变不同的K值和距离度量方式，可视化聚类结果;

以下是Python代码实现KMeans聚类，可以调整K值和距离度量方式，并可视化聚类结果：

import numpy as np
import matplotlib.pyplot as plt

# 定义计算欧式距离的函数
def euclidean_distance(x1, x2):
    return np.sqrt(np.sum((x1 - x2) ** 2))

# 定义计算曼哈顿距离的函数
def manhattan_distance(x1, x2):
    return np.sum(np.abs(x1 - x2))

class KMeans:
    def __init__(self, k=3, max_iters=100, distance='euclidean'):
        self.k = k
        self.max_iters = max_iters
        self.distance = distance
        self.clusters = [[] for _ in range(self.k)]
        self.centroids = []

    def fit(self, X):
        # 随机初始化质心
        self.centroids = [X[np.random.choice(len(X))] for _ in range(self.k)]
        
        # 迭代更新质心和聚类
        for _ in range(self.max_iters):
            # 初始化聚类
            self.clusters = [[] for _ in range(self.k)]
            
            # 计算每个样本到质心的距离，并分配到最近的聚类
            for x in X:
                distances = [euclidean_distance(x, c) if self.distance=='euclidean'
                             else manhattan_distance(x, c) for c in self.centroids]
                closest_cluster = np.argmin(distances)
                self.clusters[closest_cluster].append(x)
            
            # 计算每个聚类的新质心
            for i, cluster in enumerate(self.clusters):
                self.centroids[i] = np.mean(cluster, axis=0)

    def predict(self, X):
        # 预测每个样本所属的聚类
        distances = np.zeros((len(X), self.k))
        for i, x in enumerate(X):
            for j, c in enumerate(self.centroids):
                distances[i, j] = euclidean_distance(x, c) if self.distance=='euclidean' \
                                  else manhattan_distance(x, c)
        return np.argmin(distances, axis=1)
      
# 测试代码
X = np.array([[1, 2], [1, 4], [1, 0],
              [4, 2], [4, 4], [4, 0]])
kmeans = KMeans(k=2, max_iters=100, distance='euclidean')
kmeans.fit(X)
clusters = kmeans.predict(X)
colors = ['r', 'g', 'b', 'y', 'c', 'm']
for i, cluster in enumerate(kmeans.clusters):
    for x in cluster:
        plt.scatter(x[0], x[1], color=colors[i])
for c in kmeans.centroids:
    plt.scatter(c[0], c[1], marker='x', s=100, color='k')
plt.show()

可以调整K值和距离度量方式，例如：

kmeans = KMeans(k=3, max_iters=100, distance='manhattan')

可以看到，聚类结果会随着K值和距离度量方式的不同而发生变化。