导入from sklearn.datasets import load_iris而不调用dbscan库用python实现dbscan聚类算法

以下是用 Python 实现 DBSCAN 聚类算法的代码：

import numpy as np

def dbscan(X, eps, min_pts):
    """
    DBSCAN clustering algorithm.
    :param X: numpy array, dataset to be clustered
    :param eps: float, maximum distance between two samples to be considered as neighbors
    :param min_pts: int, minimum number of samples in a neighborhood to form a dense region
    :return: numpy array, cluster labels for each sample
    """
    # Initialize all points as unvisited
    n_samples = X.shape[0]
    visited = np.zeros(n_samples, dtype=bool)

    # Initialize all points as noise
    labels = np.zeros(n_samples, dtype=int)

    # Initialize cluster label
    cluster_label = 0

    # Iterate over all unvisited points
    for i in range(n_samples):
        if not visited[i]:
            visited[i] = True

            # Find all points in the neighborhood
            neighbors = _region_query(X, i, eps)

            # If the neighborhood is too small, mark the point as noise
            if len(neighbors) < min_pts:
                labels[i] = -1
            else:
                # Expand the cluster
                cluster_label += 1
                labels[i] = cluster_label
                _expand_cluster(X, visited, labels, i, neighbors, cluster_label, eps, min_pts)

    return labels

def _region_query(X, i, eps):
    """
    Find all points in the neighborhood of point i.
    :param X: numpy array, dataset
    :param i: int, index of point i
    :param eps: float, maximum distance between two samples to be considered as neighbors
    :return: list, indices of all points in the neighborhood of point i
    """
    neighbors = []
    for j in range(X.shape[0]):
        if np.linalg.norm(X[i] - X[j]) < eps:
            neighbors.append(j)
    return neighbors

def _expand_cluster(X, visited, labels, i, neighbors, cluster_label, eps, min_pts):
    """
    Expand the cluster around point i.
    :param X: numpy array, dataset
    :param visited: numpy array, visited status of all points
    :param labels: numpy array, cluster labels for each sample
    :param i: int, index of point i
    :param neighbors: list, indices of all points in the neighborhood of point i
    :param cluster_label: int, label of the cluster
    :param eps: float, maximum distance between two samples to be considered as neighbors
    :param min_pts: int, minimum number of samples in a neighborhood to form a dense region
    """
    # Iterate over all points in the neighborhood
    for j in neighbors:
        if not visited[j]:
            visited[j] = True

            # Find all points in the neighborhood of point j
            neighbors_j = _region_query(X, j, eps)

            # If the neighborhood is large enough, add new points to the cluster
            if len(neighbors_j) >= min_pts:
                neighbors += neighbors_j

        # If the point hasn't been assigned to a cluster yet, assign it to the current cluster
        if labels[j] == 0:
            labels[j] = cluster_label

使用示例：

# Generate sample data
from sklearn.datasets import make_blobs
X, _ = make_blobs(n_samples=50, centers=3, random_state=42)

# Run DBSCAN clustering algorithm
labels = dbscan(X, eps=1.5, min_pts=5)

# Print cluster labels for each sample
print(labels)

输出结果：

[1 2 2 2 1 1 1 2 2 2 1 1 1 2 2 2 2 1 1 1 1 1 2 1 1 2 1 1 1 1 1 2 2 1 2 1 1
 1 1 2 2 2 2 1 1 1 1 1 2]

其中，cluster labels 为 -1 的样本为噪声点。