DBSCAN implemented in scikit-learn )
#!/usr/bin/env python
# -*- coding: utf-8 -*-
import sys
import numpy as np
from sklearn import cluster
"""Specify parameters"""
dbscan = cluster.DBSCAN(eps=float(sys.argv[1]), min_samples=int(sys.argv[2]))
"""Read data"""
data_list = []
for line in open(sys.argv[3]):
x = map(float, line.rstrip().split(' '))
data_list.append(x)
data = np.array(data_list)
"""Clustering"""
dbscan.fit(data)
"""View results"""
labels = dbscan.labels_
for i in range(len(labels)):
if labels[i] != -1:
print labels[i], data[i]
Prepare the following file that describes the sample data in the row and the value of the attribute to be written in the column.
0 1
8.5 6
2 0
1.5 0
1 1.5
10 5
9 6
8 5.5
9.5 5.6
100 100
-100 -50
1 0
Execute by passing eps, min_samples, data_file in this order as arguments
>> python dbscan.py 1.5 3 data
0.0 [ 0. 1.]
1.0 [ 8.5 6. ]
0.0 [ 2. 0.]
0.0 [ 1.5 0. ]
0.0 [ 1. 1.5]
1.0 [ 10. 5.]
1.0 [ 9. 6.]
1.0 [ 8. 5.5]
1.0 [ 9.5 5.6]
0.0 [ 1. 0.]
dbscan.labels_ shows which cluster each sample was assigned to.
When it is -1, it means that the noise cannot be assigned to any cluster.
Recommended Posts