Faster solution instead of array_unique()

[rvadym]

Ticket is related to this solution

// this can be also solved with 
// return count(array_unique($A));

This is result with array_unique()
https://codility.com/demo/results/demoATDHG8-X7Z/
Only 66% for Performance

But according to this article this task can be solved with native php functions keeping good Performance score

return count(array_flip(array_flip(array_reverse($A,true))));

;)

[paljinov]

I've tested array_unique approach.
I am getting 100/100, and obviously you got 100/66, only different thing is input, which is larger in your case, so the time is slower. With growing array $A your native function use is increasingly better solution. Why this is the case, is explained in the link that you've provided.

Problematic test is large_random1 chaotic sequence of values from [-1M..1M], length=100K:

SOLUTION	Time	Link
count(array_unique($A)) - YOU	0.944 s	https://codility.com/demo/results/demoATDHG8-X7Z/
count(array_unique($A)) - ME	0.516 s	https://codility.com/demo/results/demoGCBB35-KFJ/
count(array_flip(array_flip(array_reverse($A,true)))	0.204 s	https://codility.com/demo/results/demo8C3NFT-2D5/
MY CUSTOM CODE	0.320 s	https://codility.com/demo/results/demo8CTQH9-EF6/

Please, can you push your solution, just change:

// this can be also solved with 
// return count(array_unique($A));

in

// this can be also solved with 
// return count(array_flip(array_flip(array_reverse($A, true))));