基于多个键从多维数组中删除重复项

Sorry if this was asked before, but I searched a lot and couldn't find a solution. I've been trying to solve this problem for a while now, and couldn't write the function for it.

I have an array like that:

    $numbers = array(
        array("tag" => "developer", "group" => "grp_1", "num" => "123123"),
        array("tag" => "developer", "group" => "grp_2", "num" => "111111"),
        array("tag" => "student", "group" => "grp_1", "num" => "123123"),
        array("tag" => "student", "group" => "grp_2", "num" => "123123"),
        array("tag" => "developer", "group" => "grp_3", "num" => "111111"),
    );

I need to write a function, that removes the duplicates off this array, based on multiple keys, so my function call should look something like that:

unique_by_keys($numbers, array("num","group"));

In other terms, one number can't be in the same group more than once.

After calling unique_by_keys() by array should be like that:

    $numbers = array(
        array("tag" => "developer", "group" => "grp_1", "num" => "123123"),
        array("tag" => "developer", "group" => "grp_2", "num" => "111111"),
        array("tag" => "student", "group" => "grp_2", "num" => "123123"),
        array("tag" => "developer", "group" => "grp_3", "num" => "111111"),
    );

I'd appreciate if you could help me find a solution, or lead me to the correct way of thinking. Thanks!

SOLUTION: I was able to find a solution, by writing the following function: ( I wrote it in a way that accepts many forms of $haystack arrays )

function unique_by_keys($haystack = array(), $needles = array()) {
    if (!empty($haystack) && !empty($needles)) {
        $_result = array();
        $result = array();
        $i = 0;
        foreach ($haystack as $arrayObj) {
            if (is_array($arrayObj)) {
                $searchArray = array();
                foreach ($needles as $needle) {
                    if (isset($arrayObj[$needle])) {
                        $searchArray[$needle] = $arrayObj[$needle];
                    }
                }
                if (!in_array($searchArray, $_result)) {
                    foreach ($arrayObj as $key => $value) {
                        if (in_array($key, $needles)) {
                            $_result[$i][$key] = $value;
                        }
                    }
                    $result[] = array_merge($_result[$i], $arrayObj);
                }
            } else {
                $result[] = $arrayObj;
            }
            $i++;
        }
        return $result;
    }
}

Thanks for everyone that replied!

Bhaskar's approach which assigns unique keys in the loop to remove duplicates affords a very small function for this case.

Here is a previous and unnecessarily complicated version:

function unique_by_keys($haystack=array(),$needles=array()){
    // reverse order of sub-arrays to preserve lower-indexed values
    foreach(array_reverse($haystack) as $row){
        $result[implode('',array_intersect_key($row,array_flip($needles)))]=$row; // assign unique keys
    }
    ksort($result);  // sort the sub-arrays by their assoc. keys
    return array_values($result); // replace assoc keys with indexed keys
}

This is the best/leanest solution I can come up with:

$numbers = array(
        array("tag" => "developer", "group" => "grp_1", "num" => "123123"),
        array("tag" => "developer", "group" => "grp_2", "num" => "111111"),
        array("tag" => "student",   "group" => "grp_1", "num" => "123123"),
        array("tag" => "student",   "group" => "grp_2", "num" => "123123"),
        array("tag" => "developer", "group" => "grp_3", "num" => "111111")
    );

function unique_by_keys($haystack=array(),$needles=array()){
    foreach($haystack as $row){
        $key=implode('',array_intersect_key($row,array_flip($needles)));  // declare unique key
        if(!isset($result[$key])){$result[$key]=$row;} // save row if non-duplicate
    }
    return array_values($result);
}

echo "<pre>";
var_export(unique_by_keys($numbers,array("group","num")));
echo "</pre>";

Output:

array (
  0 => 
  array (
    'tag' => 'developer',
    'group' => 'grp_1',
    'num' => '123123',
  ),
  1 => 
  array (
    'tag' => 'developer',
    'group' => 'grp_2',
    'num' => '111111',
  ),
  2 => 
  array (
    'tag' => 'student',
    'group' => 'grp_2',
    'num' => '123123',
  ),
  3 => 
  array (
    'tag' => 'developer',
    'group' => 'grp_3',
    'num' => '111111',
  ),
)

Code might not be efficient, but i will work for you :)

$result = unique_by_keys($numbers, array("num","group"));

echo "<pre>";
print_R($result);

function unique_by_keys($numbers, $arr){
    $new_array = array();
    $output = array();
    foreach ($numbers as $n){

        if(isset($new_array[$n[$arr[1]]]) && $new_array[$n[$arr[1]]] == $n[$arr[0]]){
            continue;
        }else{
            $new_array[$n[$arr[1]]] = $n[$arr[0]];
            $output[] = $n;
        }

    }
    return $output;
}

$newNumbers = array();
foreach($numbers as $key=>$values){
    $newkey = $values['group'].'__'.$values['num']; 
    $newNumbers[$newkey] = $values;
}
var_dump($newNumbers)