用python写一段密度聚类算法

下面是使用Python编写的密度聚类算法的示例代码：

import numpy as np
from sklearn.neighbors import NearestNeighbors

def density_cluster(data, eps, min_samples):
    """
    密度聚类算法
    :param data: 数据集，每行表示一个样本
    :param eps: 半径大小，用于确定一个核心点的邻域范围
    :param min_samples: 邻域内最小样本数，用于确定一个核心点
    :return: 聚类结果，每个样本的类别标签
    """
    n_samples = data.shape[0]
    visited = np.zeros(n_samples, dtype=bool)
    labels = np.zeros(n_samples, dtype=int)
    cluster_id = 1

    nbrs = NearestNeighbors(n_neighbors=min_samples).fit(data)
    distances, indices = nbrs.kneighbors(data)

    for i in range(n_samples):
        if visited[i]:
            continue

        visited[i] = True

        if len(indices[i]) < min_samples:
            labels[i] = -1  # 噪声点
        else:
            labels[i] = cluster_id

            # 扩展核心点的邻域
            neighbors = indices[i].tolist()
            j = 0
            while j < len(neighbors):
                neighbor = neighbors[j]
                if not visited[neighbor]:
                    visited[neighbor] = True
                    labels[neighbor] = cluster_id

                    if len(indices[neighbor]) >= min_samples:
                        neighbors += indices[neighbor].tolist()

                j += 1

            cluster_id += 1

    return labels

使用示例：

import numpy as np

# 生成样本数据
np.random.seed(0)
samples = np.random.randn(100, 2)

# 密度聚类
labels = density_cluster(samples, eps=0.3, min_samples=5)

# 打印聚类结果
for i, label in enumerate(labels):
    print(f"样本{i+1}的类别标签为：{label}")

这是一个简化的密度聚类算法实现，基于密度聚类的核心思想。在算法中，使用了sklearn库中的NearestNeighbors来计算样本之间的距离和邻域关系。算法会为每个样本分配一个类别标签，如果样本被认为是噪声点，则类别标签为-1