Code Coverage
 
Classes and Traits
Functions and Methods
Lines
Total
0.00%
0 / 1
0.00%
0 / 1
 CRAP
84.21%
16 / 19
DocumentLength
0.00%
0 / 1
0.00%
0 / 1
 6.14
84.21%
16 / 19
 __invoke($tfidf)
0.00%
0 / 1
 6.14
84.21%
16 / 19
<?php
/**
* This file is part of the Statistical Classifier package.
*
* (c) Cam Spiers <camspiers@gmail.com>
*
* For the full copyright and license information, please view the LICENSE
* file that was distributed with this source code.
*/
namespace Camspiers\StatisticalClassifier\Transform;
/**
* @author Cam Spiers <camspiers@gmail.com>
* @package Statistical Classifier
*/
class DocumentLength
{
public function __invoke($tfidf)
{
$transform = $tfidf;
foreach ($tfidf as $category => $documents) {
foreach ($documents as $documentIndex => $document) {
$denominator = 0;
foreach ($document as $count) {
$denominator += $count * $count;
}
$denominator = sqrt($denominator);
if ($denominator != 0) {
foreach ($document as $token => $count) {
$transform
[$category]
[$documentIndex]
[$token] = $count / $denominator;
}
} else {
throw new \RuntimeException("Cannot divide by 0 in DocumentLength transform");
}
}
}
return $transform;
}
}