# What is most efficient for GCD?

I know that Euclid's algorithm is the best algorithm for get the GCD (great common divisor) for a list the positive integer numbers. But, in the practice, you can write two codes por evaluate the gcd (for my case, i decided use java, but c/c++ may be another option).

I need to get the most efficient code of two possibilities form to programming.

Recursive Mode, you can write...

static long gcd(long a, long b){
a = Math.abs(a); b = Math.abs(b);
return (a==0)?b:gcd(b, a%b);
}


And, iterative mode, looks like ...

static long gcd(long a, long b) {
long r, i;
while(b!=0){
r = a % b;
a = b;
b = r;
}
return a;
}


Regards,

UPDATE

We can do that with the Binary GCD, and the easy code is like that

int gcd(int a, int b)
{
while(b) b ^= a ^= b ^= a %= b;
return a;
}


Great Discussion.

-
I think this is too subjective, and perhaps even better suited for StackOverflow. "Most efficient in practice" depends on many (even unpredictable) factors, such as the underlying architechture, memory hierarchy, size and form of the input etc. –  Juho Apr 22 '12 at 18:35
This is the same algorithm expressed in recursive and iterative ways. I think their difference is negligible since Euclid algorithm converges pretty fast. Choose one that fits your preference. –  pad Apr 22 '12 at 19:15
You might want to try profiling these two. Since the recursive version is a tail call, it is not unlikely that the compiler actually emits almost the same code. –  Louis Apr 22 '12 at 19:25
this is wrong. should be while b != 0, and then return a. Otherwise it bugs out on division by zero. also don't use recursion if you have really big gcds....you get a pile of stack and function states...why not just go iterative? –  Cris Stringfellow Nov 15 '12 at 12:34
Note that there are asymptotically faster GCD algorithms. E.g. en.wikipedia.org/wiki/Binary_GCD_algorithm –  Neal Young Nov 21 '12 at 17:12

As I know java doesn't supports tail recursion optimization in general, but you can test it for your case, if it doesn't supports, simple for loop should be faster else they are same as each other, but these are bit optimization, choose the one you think is easier and more readable. I should note that fastest GCD algorithm is not euclidean algorithm, Lehmer's algorithm is a bit faster.

-

Your two algorithms are equivalent (at least for positive integers, what happens with negative integers in the imperative version depends on Java's semantics for % which I don't know by heart). In the recursive version, let $a_i$ and $b_i$ be the argument of the $i$th recursive call: $$\begin{gather*} a_{i+1} = b_i \\ b_{i+1} = a_i \mathbin{\mathrm{mod}} b_i \\ \end{gather*}$$

In the imperative version, let $a'_i$ and $b'_i$ be the values of the variables a and b at the beginning of the $i$th iteration of the loop. $$\begin{gather*} a'_{i+1} = b'_i \\ b'_{i+1} = a'_i \mathbin{\mathrm{mod}} b'_i \\ \end{gather*}$$

Notice a resemblance? Your imperative version and your recursive version are calculating exactly the same values. Furthermore, they both end at the same time, when $a_i=0$ (resp. $a'_i=0$), so they perform the same number of iterations. So algorithmically speaking, there is no difference between the two. Any difference will be a matter of implementation, highly dependent on the compiler, the hardware it runs on, and quite possibly the operating system and what other programs are running concurrently.

The recursive version makes only tail recursive calls. Most compilers for imperative languages do not optimize these, and so it is likely that the code they generate will waste a little time and memory constructing a stack frame at each iteration. With a compiler that optimizes tail calls (compilers for functional languages almost always do), the generated machine code may well be the same for both (assuming you harmonize those calls to abs).

-

Just to clarify for anyone that tries to use this, the return statement for the recursive method that the OP wrote should actually be: return (b==0) ? a : gcd(b,a%b);

Otherwise you'll get a "floating point exception" (divide by zero) error.

This also assumes that a > b, which should really be checked.

-