Jaccard Coefficient in PHP

<?php

/**
 * Return the coefficient of two items based on Jaccard index
 * http://en.wikipedia.org/wiki/Jaccard_index
 *
 * Example:
 * 
 * $tags1 = "code, php, jaccard, test, items";
 * $tags2 = "test, code";
 * echo getSimilarityCoefficient( $tags1, $tags2 ); // 0.28
 *
 * $str1 = "similarity coefficient of two items";
 * $str2 = "two items are cool";
 * echo getSimilarityCoefficient( $str1, $str2, " " ); // 0.44
 *
 * @param 	string $item1 
 * @param 	string $item2 
 * @param 	string $separator 
 * @return 	float
 * @author 	Henrique Hohmann
 * @version     0.1
 */
function getSimilarityCoefficient( $item1, $item2, $separator = "," ) {
	
	$item1 = explode( $separator, $item1 );
	$item2 = explode( $separator, $item2 );
	$arr_intersection = array_intersect( $item2, $item2 );
	$arr_union = array_merge( $item1, $item2 );
	$coefficient = count( $arr_intersection ) / count( $arr_union );
	
	return $coefficient;
}
	
?>

hey sir there is a little mistake on :
$arr_intersection = array_intersect( $item2, $item2 );
change it to :
$arr_intersection = array_intersect( $item1, $item2 );
and thank you

Hello.
Thank you for sharing.

I think you should use array_unique on the return of array_merge, because array_merge adds duplicities. This way, your function is returning 0.5 for two identical sets instead of returning 1.

So, the line 29 would be:
$arr_union = array_unique(array_merge( $item1, $item2 ));

After fixing the above to errors, final function should look like this. Thanks everyone.

function getSimilarityCoefficient( $item1, $item2, $separator = "," ) {
	$item1 = explode( $separator, $item1 );
	$item2 = explode( $separator, $item2 );
	$arr_intersection = array_intersect( $item1, $item2 );
	$arr_union = array_unique(array_merge( $item1, $item2 ));
	$coefficient = count( $arr_intersection ) / count( $arr_union );
	
	return $coefficient;
}

I made some changes because similarities may consider lowercase/uppercase words the same way, check it out there: https://3v4l.org/48UZL

Here's the new code:

function getSimilarityCoefficient( $item1, $item2, $separator = "," ) {
	
	$item1 = array_unique(array_map('trim', explode( $separator, strtolower($item1) )));
	$item2 = array_unique(array_map('trim', explode( $separator, strtolower($item2) )));
    $arr_intersection = array_intersect( $item2, $item1 );
	$arr_union = array_unique(array_merge( $item1, $item2 ));
	$coefficient = count( $arr_intersection ) / count( $arr_union );
	
	return $coefficient;
}

array_intersect will return duplicates if they are present in the first array, so the results may be wrong:

echo getSimilarityCoefficient("red red red", "green blue red", " "); // 1

change line 28 to:

$arr_intersection = array_unique(array_intersect( $item1, $item2 ));