Test if number is odd or even

Question

What is the simplest most basic way to find out if a number/variable is odd or even in PHP? Is it something to do with mod?

I've tried a few scripts but.. google isn't delivering at the moment.

Solution 1

You were right in thinking mod was a good place to start. Here is an expression which will return true if $number is even, false if odd:

$number % 2 == 0

Works for every integer^PHP value, see as well Arithmetic Operators^PHP.

Example:

$number = 20;
if ($number % 2 == 0) {
  print "It's even";
}

Output:

It's even

Solution 2

Another option is a simple bit checking.

n & 1

for example:

if ( $num & 1 ) {
  //odd
} else {
  //even
}

Solution 3

Yes using the mod

$even = ($num % 2 == 0);
$odd = ($num % 2 != 0);

Solution 4

While all of the answers are good and correct, simple solution in one line is:

$check = 9;

either:

echo ($check & 1 ? 'Odd' : 'Even');

or:

echo ($check % 2 ? 'Odd' : 'Even');

works very well.

Solution 5

(bool)($number & 1)

or

(bool)(~ $number & 1)

View more solutions

Comments

[Marc B]

mod is the generic shorthand term for 'modulo', aka modular arithmetic

[ashleedawg]

More info here including: To determine odd or even it's faster and more efficient to use the bitwise & operator: $a=3; if($a&1){echo 'odd';}else{echo 'even';} //returns 'odd'

[Martijn]

If you use this in loops or large quantities, you might want to consider the bitcheck suggested by Arius2038, which is very fast. The bitcheck is my prefered method for odd/even checks.

[thomasrutter]

This would definitely be the fastest way when using integers in a language like C, by a large margin. Has anyone done benchmarks to determine if this is true also for PHP?

[Pawel Dubiel]

on my server ( 5.4.4 / cli / no opcache / i7 ) the "&" is about 10% faster then mod ( tested on array with random integer values )

[danhere]

This is a bitwise operator I believe so unless you know what you're doing with that fancyness, I would avoid this syntax.

[Rolf]

It's probably the fastest way, if the PHP engine was well coded.

[Subin]

To Downvoter : May I please know why you downvoted ?

[Robbiegod]

I'd say this is the fastest and most straight forward way. Perfect.

[snlan]

Works fine but I'm just wondering whats the logic behind this? Why is it that a value of true is given if "10 == 0" ?

[crdunst]

The logic is that any even number is divisible by 2 with no remainder. If you used $number % 10, and your $number was 20, it would return true, but not if your $number was 22. You could use $number % 10 if you were trying to find every 10th item in a loop for example.

[Hendry Tanaka]

@Tim, if 5%2=1. The logica is 2*2+1=5.How to get the 2 in php?

[Joey Sabey]

@Hendry - what are you asking exactly? How to get the quotient for a division as a whole number, or...? If that is what you mean, you just need to floor() the result; floor(5/2)=2

[grantwparks]

It's very expensive, compared to other methods.

[Subin]

@grantwparks Well, the difference between using isset & mod is only 0.5007 seconds. But, array_search is very expensive.

[grantwparks]

Those aren't seconds in the test, they're actually much less. So on the surface, it sounds reasonable. But think of this: in the results, mod came out at 2.0xxxx units of time. While isset came out at 2.5xxxx units of time. That's a 25% increase in execution time. How reasonable does it sound now ;)

[Subin]

@grantwparks Well, you're right. This answer is only for information and I suggest using mod.

[kasimir]

Might I suggest a triple = for fractional speed improvement: $number % 2 === 0

[kasimir]

Above link is dead. Here's another good explanation: catonmat.net/blog/low-level-bit-hacks-you-absolutely-must-kn‌​ow

[Subin]

@grantwparks I have update the code to include in_array which beats mod operator sometimes.

[Matthijs Wessels]

Interesting way of thinking though. It's basically the decimal version of $num & 1 :). You could also do it hexadecimal: array(0, 2, 4, 6, 8, A, C, E, F) :D.

[Jonas Lundman]

I have to admire the syntax, things that works without knowing why, gives you a reminder of how small we are in the world of fysics, math and, well, just add a row on number 1, not 2...

[GordonM]

What does this answer add that the originally accepted answer doesn't?

[GordonM]

What does this answer add that the originally accepted answer doesn't?

[Imran Azim]

it takes input from the user and take decision on that input.

[GordonM]

So nothing really.

[Martin James]

I use bitwise operators in JS quite a bit. For example if (~string.indexOf("@")) {} instead of if (string.indexOf("@") !== -1) {}. I prefer to see conditions result in a simple true or false. But yes, it can be a little confusing to people that aren't familiar with this syntax.

[ToolmakerSteve]

NOTE: to also handle float values, testing nearest integer, do if ((int)round($number) % 2 === 0). If omit "round", this is truncation, which I don't recommend: a floating point number could display as if it were a whole value (e.g. 1.0), yet be a hair below that whole value. If really need truncation, then add a small offset: if ((int)($number + 0.00001) % 2 === 0). The amount of offset to add depends on your circumstances.

[ToolmakerSteve]

Please don't do this, IMHO. Relying on the decimal string representation of an integer is not a good way to think about numeric operations. If anyone on my team used this technique, I would be quite concerned about their grasp of mathematics; suggest they speak with another programmer next time they encountered a math problem. If you want a more specific complaint, this technique doesn't generalize to other divisors: Suppose we next need to know whether a value is divisible by 3. You've gone down a dead-end road.

[ToolmakerSteve]

The is_int approach "smells" to me. It relies on the exact implementation details of integer division. I would avoid it, regardless of performance.

[ToolmakerSteve]

The is_int approach "smells" to me. It relies on the exact implementation details of integer division. I would avoid it, regardless of performance.

[ToolmakerSteve]

CAUTION: In php, the "looseness" of the language means one often encounters an integer represented as a string (which of course fails is_int test). For example, when interacting with SQL on a website. I would instead use is_numeric, which will reject null and empty string. However, that will allow floats and float-representation-strings, so may need additional tests to be thorough.

[ToolmakerSteve]

This is a nice demonstration of using case statement with the mod test.

[ToolmakerSteve]

@MartinJames: re "I prefer to see conditions result in a simple true or false." which is exactly what techniques such as !== -1 or === 0 do. Here's the problem with using bitwise operators to do anything other than a bitwise operation: you are placing a burden on the reader, to understand your intent. At minimum, you should add a comment anywhere you use that technique. Or write a well-named function and call it. Smells like an unnecessary micro-optimization to me. Seriously, if I was working with you, I would ask you to change to standard usage of obvious operators and expressions.

[Loek Bergman]

This answer is returned with the question for a basic answer in mind. You are absolutely right that in a normal application extra code is required, but that is out of scope of the question. My main point in this answer is that the operator === should be used instead of the operator ==. The last operator will return 'even' when the input is 0, "", null or false.

[ToolmakerSteve]

OK. Suggesting is_int is good. In production code I might expand this to if (!is_numeric($pArg)) ..throw-some-exception..; $p = (int)$pArg; return ($p % 2) == 0; Bug: you omitted == 0 from last snippet: return $pArg%2; returns 0 (so "false") for even numbers. Minor nit: You use === in a place where it is not at all needed. is_int can only return true or false, so === true can be safely omitted there.

[Loek Bergman]

Suggestion: Exception throwing is very important in constructing robust code, I agree. Yet in some functions I wonder if it is wise to throw an exception and return a false instead. Throwing an exception implies that the process ends. Are mistakes in all functions really that important? Bug: Could be a bug if I test with ==, yet I always test with ===. Then is 0 different from false. Minor nit: Correct that it can be omitted. I write it in PHP though, to show the other programmers that this is the check to be executed. It is for maintenance purposes only.

[Loek Bergman]

The code returns a value. In this case 0, 1 or false. The caller of this function has to test using the === in order to make this distinction. It would be more robust code indeed when the function returns 2,3 or false like this 'return ($pArg % 2) === 0 ? 2 : 3;'. Even if the caller uses == then will these values never create any confusion.

[Loek Bergman]

The function answers two questions: is it an int? If so, is it odd or even? Hence is a tri-state value required. With your expert knowledge you should have known. I programmed in Java and had to get used to the fact that 0 is not automatically different from false. I use true/false for primitive values and null for scalar ones. And always the === - operator. In Java null is an object, so I don't use it when testing primitive values.

[ToolmakerSteve]

Thank you for editing the answer to clarify the values returned by your function.

[Quentin]

That's a ludicrously inefficient approach to solving the problem, so much so that it will fall over on large numbers as it will run out of memory.