PHP : Remove duplicate values from associative array and return an associative array containing the duplicate values

Try:

Use array_filter() to get all duplicate values from array

Use array_diff() to get all unique values from array

$array = array(1=>0, 2=>1, 3=>1, 4=>2);
$counts = array_count_values($array);
$duplicates = array_filter($array, function ($value) use ($counts) {
    return $counts[$value] > 1;
});
print '<pre>';print_r($duplicates);

$result=array_diff($array,$duplicates);
print '<pre>';print_r($result);

Output:

Array
(
    [2] => 1
    [3] => 1
)

Array
(
    [1] => 0
    [4] => 2
)

The user comments on the array_unique page do shed some light on this. You will most likely find some hidden gems in those comments - its a very handy documentation.

Just a quick browser through revealed the following to remove duplicates from a multi dimensional array:

<?php
function super_unique($array)
{
  $result = array_map("unserialize", array_unique(array_map("serialize", $array)));

  foreach ($result as $key => $value)
  {
    if ( is_array($value) )
    {
      $result[$key] = super_unique($value);
    }
  }

  return $result;
}
?>

$value = reset($arr);
$key = key($arr);

(in that order)

See reset()^{PHP Manual}, key()^{PHP Manual}.

unset($arr[$key]); # in case you want to remove it.

However array_pop()^{PHP Manual} is working with the last element:

$value = end($arr);
$key = key($arr);
unset($arr[$key]); # in case you want to remove it.

See end()^{PHP Manual}.

For the fun:

list($value, $key) = array(end($arr), key($arr));

or

extract(array('value'=>end($arr), 'key'=>key($arr)));

or

end($arr);
list($key, $value) = each($arr);

or whatever style of play you like ;)

Dealing with empty arrays

It was missing so far to deal with empty arrays. So it's a need to check if there is a last (first) element and if not, set the $key to null (as null can not be an array key):

for($key=null;$key===null&&false!==$value=end($arr);)
    unset($arr[$key=key($arr)]);

This will give for a filled array like $arr = array('first' => '1st', 'last' => '2nd.');:

string(4) "2nd." # value
string(4) "last" # key
array(1) { # leftover array
  ["first"]=>
  string(3) "1st"
}

And an empty array:

bool(false) # value
NULL # key
array(0) { # leftover array
}

Afraid of using unset?

In case you don't trust unset() having the performance you need (of which I don't think it's really an issue, albeit I haven't run any metrics), you can use the native array_pop() implementation as well (but I really think that unset() as a language construct might be even faster):

end($arr);
$key = key($arr);
$value = array_pop($arr);