README

频繁模式增长算法的PHP实现

入门指南

您可以使用composer安装此包

composer require enzomc/php-fpgrowth

用法

运行算法

use EnzoMC\PhpFPGrowth\FPGrowth;

$support = 3;
$confidence = 0.7;

$fpgrowth = new FPGrowth($support, $confidence);

$transactions = [
    ['M', 'O', 'N', 'K', 'E', 'Y'],
    ['D', 'O', 'N', 'K', 'E', 'Y'],
    ['M', 'A', 'K', 'E'],
    ['M', 'U', 'C', 'K', 'Y'],
    ['C', 'O', 'O', 'K', 'I', 'E']
];

$fpgrowth->run($transactions);

$patterns = $fpgrowth->getPatterns();
$rules = $fpgrowth->getRules();

返回结果

模式以数组数组的形式返回，例如

[
    ['ITEM_1,ITEM_2' => 3],
    ['ITEM_3' => 5],
    ...
]

其中键是项目集，值是该项目集的支持度

规则以数组数组的形式返回，例如

[
    ['ITEM_1,ITEM_2', 'ITEM_3', 0.7],
    ['ITEM_4','ITEM_5', 0.7],
    ...
]

其中第一个值是规则左侧路径，第二个值是规则右侧路径，第三个值是规则的置信度

包含示例交易的输出结果

var_dump($patterns);

array(10) {
  ["M"]=>
  int(3)
  ["K,M"]=>
  int(3)
  ["Y"]=>
  int(3)
  ["K,Y"]=>
  int(3)
  ["E,O"]=>
  int(4)
  ["E,K"]=>
  int(4)
  ["E,K,O"]=>
  int(4)
  ["O"]=>
  int(4)
  ["K,O"]=>
  int(4)
  ["K"]=>
  int(5)
}

var_dump($rules);

array(11) {
  [0]=>
  array(3) {
    [0]=>
    string(1) "M"
    [1]=>
    string(1) "K"
    [2]=>
    float(1)
  }
  [1]=>
  array(3) {
    [0]=>
    string(1) "Y"
    [1]=>
    string(1) "K"
    [2]=>
    float(1)
  }
  [2]=>
  array(3) {
    [0]=>
    string(1) "O"
    [1]=>
    string(1) "E"
    [2]=>
    float(1)
  }
  [3]=>
  array(3) {
    [0]=>
    string(1) "K"
    [1]=>
    string(1) "E"
    [2]=>
    float(0.8)
  }
  [4]=>
  array(3) {
    [0]=>
    string(1) "K"
    [1]=>
    string(3) "E,O"
    [2]=>
    float(0.8)
  }
  [5]=>
  array(3) {
    [0]=>
    string(1) "O"
    [1]=>
    string(3) "E,K"
    [2]=>
    float(1)
  }
  [6]=>
  array(3) {
    [0]=>
    string(3) "E,K"
    [1]=>
    string(1) "O"
    [2]=>
    float(1)
  }
  [7]=>
  array(3) {
    [0]=>
    string(3) "E,O"
    [1]=>
    string(1) "K"
    [2]=>
    float(1)
  }
  [8]=>
  array(3) {
    [0]=>
    string(3) "K,O"
    [1]=>
    string(1) "E"
    [2]=>
    float(1)
  }
  [9]=>
  array(3) {
    [0]=>
    string(1) "K"
    [1]=>
    string(1) "O"
    [2]=>
    float(0.8)
  }
  [10]=>
  array(3) {
    [0]=>
    string(1) "O"
    [1]=>
    string(1) "K"
    [2]=>
    float(1)
  }
}