optics聚类算法python代码

好的，以下是一个简单的Python实现的optics聚类算法代码：

import numpy as np
from sklearn.neighbors import NearestNeighbors

class OPTICS:
    def __init__(self, eps, min_samples):
        self.eps = eps
        self.min_samples = min_samples

    def fit(self, X):
        self.X = X
        self.n_samples = X.shape[0]
        self.visited_ = np.zeros(self.n_samples, dtype=bool)
        self.ordering_ = []
        self.core_distances_ = np.zeros(self.n_samples)
        self.reachability_distances_ = np.ones(self.n_samples) * np.inf
        self.process()

    def process(self):
        self.tree_ = NearestNeighbors(radius=self.eps, algorithm='ball_tree').fit(self.X)
        for point_idx in range(self.n_samples):
            if not self.visited_[point_idx]:
                self.visited_[point_idx] = True
                self.ordering_.append(point_idx)
                neighbors_idx = self.tree_.radius_neighbors([self.X[point_idx]], return_distance=False)[0]
                n_neighbors = len(neighbors_idx)
                if n_neighbors >= self.min_samples:
                    self.core_distances_[point_idx] = np.max(self.tree_.kneighbors([self.X[point_idx]], n_neighbors=self.min_samples)[0])
                    self.update(neighbors_idx, point_idx)

        self.ordering_ = np.array(self.ordering_)
        self.cluster_ordering_ = np.argsort(self.reachability_distances_)
        self.reachability_distances_[self.reachability_distances_ == np.inf] = -1

    def update(self, neighbors_idx, point_idx):
        core_dist = self.core_distances_[point_idx]
        for neighbor in neighbors_idx:
            if not self.visited_[neighbor]:
                new_reach_dist = max(core_dist, np.linalg.norm(self.X[point_idx] - self.X[neighbor]))
                if np.isinf(self.reachability_distances_[neighbor]):
                    self.reachability_distances_[neighbor] = new_reach_dist
                else:
                    self.reachability_distances_[neighbor] = min(new_reach_dist, self.reachability_distances_[neighbor])

    def extract_clusters(self):
        clusters = []
        idx = 0
        while idx < self.n_samples:
            if self.reachability_distances_[self.cluster_ordering_[idx]] == -1:
                idx += 1
                continue

            cluster = [self.cluster_ordering_[idx]]
            while True:
                idx += 1
                if idx == self.n_samples:
                    break
                if self.reachability_distances_[self.cluster_ordering_[idx]] == -1:
                    continue
                if self.reachability_distances_[self.cluster_ordering_[idx]] > self.eps:
                    break
                cluster.append(self.cluster_ordering_[idx])
            clusters.append(cluster)

        return clusters