TfIdfTransformer::countTokensFrequency() - Code Metrics - Inspection of "Tweaks to the TF-LDF transformer" - php-ai/php-ml - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#392)

unknown

created 2019-07-10 16:43 UTC

TfIdfTransformer::countTokensFrequency() A

↳ Parent: TfIdfTransformer

Complexity

Conditions	4
Paths	4

Size

Total Lines	8
Code Lines	5

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
eloc	5
dl	0
loc	8
rs	10
c	0
b	0
f	0
cc	4
nc	4
nop	1

<?php

declare(strict_types=1);

namespace Phpml\FeatureExtraction;

use Phpml\Transformer;

class TfIdfTransformer implements Transformer
{
    /**
     * @var array
     */
    private $idf = [];

    /**
     * The term counts.
     *
     * @var array
     */
    private $termCounts = [];

    /**
     * The minimum accepted term frequency.
     *
     * @var int
     */
    private $minTf;

    /**
     * The minimum accepted IDF value.
     *
     * @var float
     */
    private $minIdf;

    public function __construct(array $samples = [], int $minTf = 0, float $minIdf = 0.0)
    {
        if (count($samples) > 0) {
            $this->fit($samples);
        }
        $this->minTf = $minTf;
        $this->minIdf = $minIdf;
    }

    public function fit(array $samples, ?array $targets = null): void
    {
        $this->termCounts = array_fill_keys(array_keys($samples[0]), 0);

        foreach ($samples as $sample) {
            foreach ($sample as $index => $count) {
                if ($count > 0) {
                    $this->termCounts[$index]++;
                }
            }
        }

        $count = count($samples);
        $this->idf = array_map(
            function (float $value) use ($count): float {
                return $value > 0.0
                    ? log($count / $value, 10)
                    : 0;
            },
            $this->termCounts
        );
    }

    public function transform(array &$samples, ?array &$targets = null): void
    {
        array_walk($samples, function (array &$sample): void {
            foreach ($sample as $index => &$feature) {
                if ($this->termCounts[$index] < $this->minTf || $this->idf[$index] < $this->minIdf) {
                    unset($sample[$index]);

                    continue;
                }
                $feature *= $this->idf[$index];
            }
            $sample = array_values($sample);
        });
    }
}


1			<?php
2
3			declare(strict_types=1);
4
5			namespace Phpml\FeatureExtraction;
6
7			use Phpml\Transformer;
8
9			class TfIdfTransformer implements Transformer
10			{
11			/**
12			* @var array
13			*/
14			private $idf = [];
15
16			/**
17			* The term counts.
18			*
19			* @var array
20			*/
21			private $termCounts = [];
22
23			/**
24			* The minimum accepted term frequency.
25			*
26			* @var int
27			*/
28			private $minTf;
29
30			/**
31			* The minimum accepted IDF value.
32			*
33			* @var float
34			*/
35			private $minIdf;
36
37			public function __construct(array $samples = [], int $minTf = 0, float $minIdf = 0.0)
38			{
39			if (count($samples) > 0) {
40			$this->fit($samples);
41			}
42			$this->minTf = $minTf;
43			$this->minIdf = $minIdf;
44			}
45
46			public function fit(array $samples, ?array $targets = null): void
47			{
48			$this->termCounts = array_fill_keys(array_keys($samples[0]), 0);
49
50			foreach ($samples as $sample) {
51			foreach ($sample as $index => $count) {
52			if ($count > 0) {
53			$this->termCounts[$index]++;
54			}
55			}
56			}
57
58			$count = count($samples);
59			$this->idf = array_map(
60			function (float $value) use ($count): float {
61			return $value > 0.0
62			? log($count / $value, 10)
63			: 0;
64			},
65			$this->termCounts
66			);
67			}
68
69			public function transform(array &$samples, ?array &$targets = null): void
70			{
71			array_walk($samples, function (array &$sample): void {
72			foreach ($sample as $index => &$feature) {
73			if ($this->termCounts[$index] < $this->minTf \|\| $this->idf[$index] < $this->minIdf) {
74			unset($sample[$index]);
75
76			continue;
77			}
78			$feature *= $this->idf[$index];
79			}
80			$sample = array_values($sample);
81			});
82			}
83			}
84

php-ai / php-ml

Pull Request — master (#392)

TfIdfTransformer::countTokensFrequency() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like