Question

I have an array like this:

$elements = array(
    array("name" => "Paco", "group" => 2),
    array("name" => "Juan", "group" => 3),
    array("name" => "Marta", "group" => 2),
    array("name" => "Julia", "group" => 4),
    array("name" => "Silvia", "group" => 2)
);

I need to avoid repeating the group. If there are several elements with the same group, the final array must have only the first of each group:

$elements = array(
    array("name" => "Paco", "group" => 2),
    array("name" => "Juan", "group" => 3),
    array("name" => "Julia", "group" => 4)
);

Is there a better way than foreaching the array and processing each row? Something like GROUP BY in SQL?

Answer 1

Seeing as a lot of strange code has come up, some are wildly inefficient or just strange in my opinion:

$values = array(); 

foreach($elements as $element) { 
    if (!isset($values[$element['group']])) {
        $values[$element['group']] = $element; 
    }
} 

And what you should have now is an array where only the FIRST occurence of a group is saved and only requires one pass, O(N). If you don’t want the index to be the group id then just do:

$values2 = array_values($values);

PS. If you really dislike array_values(), then you could also do it like this (or some variation of this):

$values = array();
$groupkeys = array();

foreach($elements as $element) { 
    if (!isset($groupkeys[$element['group']])) {
         $values[] = $element;
         $groupkeys[$element['group']] = TRUE;
    } 
}