K中心点算法（k-mediods）python代码

以下是使用Python实现K中心点算法（k-mediods）的代码示例：

import numpy as np

def k_medoids(X, k):
    n_samples, n_features = X.shape

    # Initialize medoids randomly
    medoids = np.random.choice(n_samples, size=k, replace=False)
    
    # Assign each sample to the closest medoid
    labels = np.zeros(n_samples)
    for i in range(n_samples):
        distances = np.sum((X[i] - X[medoids])**2, axis=1)
        labels[i] = np.argmin(distances)

    # Update medoids as the sample that minimizes the sum of distances
    for i in range(k):
        indices = np.where(labels == i)[0]
        distances = np.sum((X[indices][:, np.newaxis, :] - X[np.newaxis, medoids])**2, axis=2)
        new_medoid = indices[np.argmin(np.sum(distances, axis=0))]
        medoids[i] = new_medoid

    # Assign each sample to the closest medoid (again)
    labels = np.zeros(n_samples)
    for i in range(n_samples):
        distances = np.sum((X[i] - X[medoids])**2, axis=1)
        labels[i] = np.argmin(distances)

    return medoids, labels

其中，输入参数X是一个numpy数组，表示待聚类的数据集，k是指定的聚类个数。函数返回medoids和labels，分别表示聚类的中心点和每个数据点所属的簇标签。