|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare (strict_types = 1); |
|
4
|
|
|
|
|
5
|
|
|
namespace Phpml\Clustering; |
|
6
|
|
|
|
|
7
|
|
|
use Phpml\Math\Distance; |
|
8
|
|
|
use Phpml\Math\Distance\Euclidean; |
|
9
|
|
|
|
|
10
|
|
|
class DBSCAN implements Clusterer |
|
11
|
|
|
{ |
|
12
|
|
|
/** |
|
13
|
|
|
* @var float |
|
14
|
|
|
*/ |
|
15
|
|
|
private $epsilon; |
|
16
|
|
|
|
|
17
|
|
|
/** |
|
18
|
|
|
* @var int |
|
19
|
|
|
*/ |
|
20
|
|
|
private $minSamples; |
|
21
|
|
|
|
|
22
|
|
|
/** |
|
23
|
|
|
* @var Distance |
|
24
|
|
|
*/ |
|
25
|
|
|
private $distanceMetric; |
|
26
|
|
|
|
|
27
|
|
|
/** |
|
28
|
|
|
* @param float $epsilon |
|
29
|
|
|
* @param int $minSamples |
|
30
|
|
|
* @param Distance $distanceMetric |
|
31
|
|
|
*/ |
|
32
|
|
|
public function __construct($epsilon = 0.5, $minSamples = 3, Distance $distanceMetric = null) |
|
33
|
|
|
{ |
|
34
|
|
|
if (null === $distanceMetric) { |
|
35
|
|
|
$distanceMetric = new Euclidean(); |
|
36
|
|
|
} |
|
37
|
|
|
|
|
38
|
|
|
$this->epsilon = $epsilon; |
|
39
|
|
|
$this->minSamples = $minSamples; |
|
40
|
|
|
$this->distanceMetric = $distanceMetric; |
|
41
|
|
|
} |
|
42
|
|
|
|
|
43
|
|
|
/** |
|
44
|
|
|
* @param array $samples |
|
45
|
|
|
* |
|
46
|
|
|
* @return array |
|
47
|
|
|
*/ |
|
48
|
|
|
public function cluster(array $samples) |
|
49
|
|
|
{ |
|
50
|
|
|
$clusters = []; |
|
51
|
|
|
$visited = []; |
|
52
|
|
|
|
|
53
|
|
|
foreach ($samples as $index => $sample) { |
|
54
|
|
|
if (isset($visited[$index])) { |
|
55
|
|
|
continue; |
|
56
|
|
|
} |
|
57
|
|
|
$visited[$index] = true; |
|
58
|
|
|
|
|
59
|
|
|
$regionSamples = $this->getSamplesInRegion($sample, $samples); |
|
60
|
|
|
if (count($regionSamples) >= $this->minSamples) { |
|
61
|
|
|
$clusters[] = $this->expandCluster($regionSamples, $visited); |
|
62
|
|
|
} |
|
63
|
|
|
} |
|
64
|
|
|
|
|
65
|
|
|
return $clusters; |
|
66
|
|
|
} |
|
67
|
|
|
|
|
68
|
|
|
/** |
|
69
|
|
|
* @param array $localSample |
|
70
|
|
|
* @param array $samples |
|
71
|
|
|
* |
|
72
|
|
|
* @return array |
|
73
|
|
|
*/ |
|
74
|
|
|
private function getSamplesInRegion($localSample, $samples) |
|
75
|
|
|
{ |
|
76
|
|
|
$region = []; |
|
77
|
|
|
|
|
78
|
|
|
foreach ($samples as $index => $sample) { |
|
79
|
|
|
if ($this->distanceMetric->distance($localSample, $sample) < $this->epsilon) { |
|
80
|
|
|
$region[$index] = $sample; |
|
81
|
|
|
} |
|
82
|
|
|
} |
|
83
|
|
|
|
|
84
|
|
|
return $region; |
|
85
|
|
|
} |
|
86
|
|
|
|
|
87
|
|
|
/** |
|
88
|
|
|
* @param array $samples |
|
89
|
|
|
* @param array $visited |
|
90
|
|
|
* |
|
91
|
|
|
* @return array |
|
92
|
|
|
*/ |
|
93
|
|
|
private function expandCluster($samples, &$visited) |
|
94
|
|
|
{ |
|
95
|
|
|
$cluster = []; |
|
96
|
|
|
|
|
97
|
|
|
foreach ($samples as $index => $sample) { |
|
98
|
|
|
if (!isset($visited[$index])) { |
|
99
|
|
|
$visited[$index] = true; |
|
100
|
|
|
$regionSamples = $this->getSamplesInRegion($sample, $samples); |
|
101
|
|
|
if (count($regionSamples) > $this->minSamples) { |
|
102
|
|
|
$cluster = array_merge($regionSamples, $cluster); |
|
103
|
|
|
} |
|
104
|
|
|
} |
|
105
|
|
|
|
|
106
|
|
|
$cluster[] = $sample; |
|
107
|
|
|
} |
|
108
|
|
|
|
|
109
|
|
|
return $cluster; |
|
110
|
|
|
} |
|
111
|
|
|
} |
|
112
|
|
|
|
php-ai /
php-ml
