[f.atyeh]

I have some files and I want to group them. these files include the name of images. I count the same images and sort them descending and chose 10 images with highest rate . so final files include 10 images and their rates. For example :
file f1: image a1 wirh rate m1
image a2 with rate m2 , ...
file f2 : image a2 with rate n1
image a3 with rate n2 , ...
and other files ...
Kind of images are unlimited.
The rate and kind of images both are important for me. now I want to find same files based on these 2 items.
These files may have overlap too .for example f1 , f2 , f9 are in the same group, f2 , f4 , f5 are in the same. f2 has overlap.
How can I group them in your idea.
Thanks,

[8pla.net]

Worth consideration may be multidimensional arrays which contain one or more arrays. 

Pseudocode version 0.1:

Code

$image=array (
  0 => 'a1',
  1 => 'a2',
  2 => 'a3',
  3 => 'a4',
);

$rate=array (
  0 => 'm1',
  1 => 'm2',
  2 => 'n1',
  3 => 'n2',
);

$file=array (
  0 => 
  array (
    0 => 'a1',
    1 => 'm1',
  ),
  1 => 
  array (
    0 => 'a2',
    1 => 'm2',
  ),
  2 => 
  array (
    0 => 'a3',
    1 => 'n1',
  ),
  3 => 
  array (
    0 => 'a4',
    1 => 'n2',
  ),
);

This may be one of the simplest approaches to consider.  And it can be further customized to the exact requirement.  But I think is should be part of the discussion.   Then we have consideration of the more advanced  possibilities to discuss, which of course becomes more exciting.