用DBSCAN聚类经纬度坐标

最新推荐文章于 2026-02-25 06:03:38 发布

原创最新推荐文章于 2026-02-25 06:03:38 发布 · 9.7k 阅读

81 ·

本内容遵循CC 4.0 BY-SA版权协议

收录于

数据挖掘

本文介绍了一种基于密度的空间聚类算法DBSCAN的实现与应用，通过使用Python的sklearn库，该算法能有效识别地理坐标数据中的高密度区域，形成聚类并计算出每个聚类的中心点，适用于GPS数据处理与分析。

用基于密度的聚类算法，计算坐标点聚集地，很好用。

import pandas as pd
import numpy as np
from sklearn.cluster import DBSCAN
from sklearn import metrics
from sklearn.cluster import KMeans
import os

def dbscan(input_file):
    ## 纬度在前，经度在后 [latitude, longitude]
    columns=['lat','lon'] 
    in_df = pd.read_csv(input_file, sep=',', header=None, names=columns)

    #represent GPS points as (lat, lon)
    coords = in_df.as_matrix(columns=['lat','lon'])

    #earth's radius in km
    kms_per_radian = 6371.0086
    #define epsilon as 0.5 kilometers, converted to radians for use by haversine
    #This uses the 'haversine' formula to calculate the great-circle distance between two points
    # that is, the shortest distance over the earth's surface
    # http://www.movable-type.co.uk/scripts/latlong.html
    epsilon = 0.5 / kms_per_radian

    # radians() Convert angles from degrees to radians
    db = DBSCAN(eps=epsilon, min_samples=15, algorithm='ball_tree', metric='haversine').fit(np.radians(coords))
    cluster_labels = db.labels_

    # get the number of clusters (ignore noisy samples which are given the label -1)
    num_clusters = len(set(cluster_labels) - set([-1]))

    print( 'Clustered ' + str(len(in_df)) + ' points to ' + str(num_clusters) + ' clusters')

    # turn the clusters in to a pandas series
    #clusters = pd.Series([coords[cluster_labels == n] for n in range(num_clusters)])
    #print(clusters)
    kmeans = KMeans(n_clusters=1, n_init=1, max_iter=20, random_state=20)

    for n in range(num_clusters):
        # print('Cluster ', n, ' all samples:')
        one_cluster = coords[cluster_labels == n]
        #print(one_cluster[:1])
        #clist = one_cluster.tolist()
        #print(clist[0])
        kk = kmeans.fit(one_cluster)
        print(kk.cluster_centers_)

def main():
    path = './datas'
    filelist = os.listdir(path)
    for f in filelist:
        datafile = os.path.join(path,f)
        print(datafile)
        dbscan(datafile)

if __name__ == '__main__':
    main()

下面几个博客，写的很好：

https://www.cnblogs.com/pinard/p/6208966.html

http://shiyanjun.cn/archives/1288.html

https://www.biaodianfu.com/dbscan.html

标签

#聚类算法