remove duplicates and move those that had duplicates to beginning of multi-dimensional array php

Question

I'm creating a search function for a csv. With the result of my search I then want to remove duplicates and then move those that had duplicates to the beginning of the array.

  Array
    (
        [0] => Array
        (
            [0] => COD 21
            [1] => 4567
            [2] => Robert Wahl
        )
    [1] => Array
        (
            [0] => RT 5
            [1] => 1234
            [2] => Robert Merrill
        )
    [2] => Array
        (
            [0] => XD 62
            [1] => 1653
            [2] => Robert Hill
        )
    [3] => Array
        (
            [0] => RT 5
            [1] => 1234
            [2] => Robert Merrill
        )
)

I have a function that removes duplicates, but I'm not sure how to get it to move the values that have duplicates to the beginning of the array, and order them according to those with the most duplicates:

function arrayUnique($array) { 
    $aux_ini=array(); 
    $result=array(); 
    for($n=0;$n<count($array);$n++) 
    { 
        $aux_ini[]=serialize($array[$n]); 
    } 
    $mat=array_unique($aux_ini); 
    for($n=0;$n<count($array);$n++) 
    { 
        $result[]=unserialize($mat[$n]); 
    } 
    return $result; 
}

Any and all help is appreciated.

Very good idea to serialize array. It give ability to use many functions with this array — splash58, Jul 09 '15 at 16:06

score 3 · Accepted Answer · answered Jul 09 '15 at 15:54

3

function arrayUnique($array) { 
    // count values
    $arr = array_count_values (array_map('serialize',$array)); 
    // sort in desc order
    arsort($arr);
    // take keys and restore array
    return array_map('unserialize',array_keys($arr));
}

answered Jul 09 '15 at 15:54

splash58

26,043
3
22
34

`array_count_values` kind of makes my suggestion laughable! – Steve Jul 09 '15 at 15:59
@Steve yes, so a little easy – splash58 Jul 09 '15 at 16:07
Thanks splash58! I now have a greater understanding of array_map and array_count_values, and how I can utilize those in the future. – Todd Austin Jul 09 '15 at 16:39

Steve · Answer 2 · 2015-07-09T16:01:38.277

2

function array_unique_ordered_by_duplicates($array)
{
  //serialize inner arrays
  $serialized = array_map('serialize', $array);
  //create an array indexed by the serialized values, with all keys having a value of 0
  $flipped = array_map(function($el){ 
        return 0; 
      },array_flip($serialized));
  //increment the count each time an element is encountered, thereby gicing duplicates a higher value
  foreach($serialized as $v){
    $flipped[$v]+=1;
  }
  //sort descending on the calculated value, putting the most duplicated items first
  arsort($flipped);
  //return unserialized, deduped, orded array
  return array_map('unserialize',array_keys($flipped));

}

Note that splash58's is a more efficient version of the above, using php's array_count_values, and should be used instead

edited Jul 09 '15 at 16:01

answered Jul 09 '15 at 15:55

Steve

20,703
5
41
67

Awesome! This works perfectly. Now to step through it on my own so I can understand how this works. Thanks Steve. – Todd Austin Jul 09 '15 at 16:10
I'm going to go ahead and use splash58's answer, thanks for yours as well. – Todd Austin Jul 09 '15 at 16:29
@Steve I am not guilty :)) – splash58 Jul 09 '15 at 16:33

score 1 · Answer 3 · answered Jul 09 '15 at 15:58

A little bit funky but I think the code is relatively easy to follow.

$csv = array(array('COD 21','4567','Robert Wahl')
            ,array('RT 5','1234','Robert Merrill')
            ,array('XD 62','1653','Robert Hill')
            ,array('RT 5','1234','Robert Merrill')
            );

echo "Begin...<br>\n";

$serials = array();
$dupeqty = array();
foreach ($csv as $csvkey => $csvdata)
{
  $dupekey = serialize($csvdata);

  if ( !isset($serials[$dupekey]) )
    $serials[$dupekey] = $csvkey;

  if ( !isset($dupeqty[$dupekey]) )
    $dupeqty[$dupekey] = 0;
  $dupeqty[$dupekey]++;
}

arsort($dupeqty);
foreach ($dupeqty as $dupekey => $times)
  echo $dupekey . " occurs $times times<br>\n";

flush();

echo "<br>\n";
echo "Accessing the original array...<br>\n";
foreach ($dupeqty as $dupekey => $times)
{
  $data = $csv[$serials[$dupekey]];
  echo '... data ' . $data[0] . ' ' . $data[1] . ' ' . $data[2] . "<br>\n";
}

Running it generates:

Begin...
a:3:{i:0;s:4:"RT 5";i:1;s:4:"1234";i:2;s:14:"Robert Merrill";} occurs 2 times
a:3:{i:0;s:5:"XD 62";i:1;s:4:"1653";i:2;s:11:"Robert Hill";} occurs 1 times
a:3:{i:0;s:6:"COD 21";i:1;s:4:"4567";i:2;s:11:"Robert Wahl";} occurs 1 times

Accessing the original array...
... data RT 5 1234 Robert Merrill
... data XD 62 1653 Robert Hill
... data COD 21 4567 Robert Wahl

this looks a lot like what I started out coding, but for some reason I ran into issues if there were more than 2 matching entries. — Todd Austin, Jul 09 '15 at 16:41
To be sure I have tested it and it does work with more than two matching entries. The concise answer is obviously the best but perhaps some will find this one illustrative. — A Smith, Jul 09 '15 at 17:06

score 0 · Answer 4 · answered Jul 09 '15 at 15:30

0

try this,

function array_multidimensional($array)
{
$serialized = array_map('serialize', $array);
$array = array_unique($serialized);
return array_intersect_key($array, $unique);
}

answered Jul 09 '15 at 15:30

remove duplicates and move those that had duplicates to beginning of multi-dimensional array php

4 Answers4