Thread: Card shuffle game

There is something called the "Fisher-Yates" algorithm that you can google. It has some variations. Here's my example:

Code:

#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <time.h>

void FYshuffle (int *ray, int len) {
	int i, j, tmp, x;
	srand(time(0));
	for (i=len-1;i>1;i--) {
		x = RAND_MAX/(i-1); 
		j=rand()/x;
		printf("%d Max: %d Actual: %d\n",i,RAND_MAX/x, j);
		if (j==i) continue;
		tmp = ray[i];
		ray[i] = ray[j];
		ray[j] = tmp;
	}
}

int main(void) {
	int ray[10] = {0,1,2,3,4,5,6,7,8,9}, i;
	FYshuffle(ray,10);
	for (i=0;i<10;i++) printf("%d\n",ray[i]);
	return 0;
}

If you are like me, you will find all the examples confusing until you write one yourself, and realize it looks like one of the examples

The idea is to count backward thru the array and swap with a preceding element. You could also count forward and swap with a subsequent element I guess -- I think the calculations/necessary operations are actually simpler this way.

Of course, probably just counting thru and swapping with any other element would be random too...

Originally Posted by sean

Ha ha - I was thinking, "Fisher and Yates"? That's the Knuth algorithm! Turns out they're the same - you learn something new everyday!

I think if you google "algorithm Knuth" you will simply get a list of every other C code description on the web, so maybe that is just as well

ps. my example can be optimized by removing some variables but it is probably easier to understand this way, I hope, WRT using rand().