Thread: Find minimum in an array, using recurssion.

This seemingly right code, don't know why gives 0 always as output. Please, help debugging it...

Code:

#include <stdio.h>

int min(int *arr,int size)
{
	static int *m;

	if(!m) m=arr;

	return size ? ( *m>*++arr ? min((m=arr),size-1) : min(arr,size-1) ) : *m;
}

int main(void)
{
	int i,size;

	printf("Enter array size:");
	scanf("%d",&size);

	int arr[size];

	printf("Enter array elements:");
	for(i=0;i<size;i++)
	scanf("%d", arr+i);

	printf("The minimum no. is %d",min(arr,size));

	return 0;
}

>> min((m=arr),size-1)

The result of an assignment expression is the left hand side. Are you sure you want to pass in m and not arr?

Also I wouldn't use a static variable at all. It's just far too much trouble -- min can only really be used with one array because they can only be initialized once and your code to assign to m is no help.

Yeah, because its equivalent whether I pass m or arr, its essentially the same address.

It may be the same address, but the variables clearly don't mean the same. m is supposed to point at the minimum element. You start using m to iterate through the array and the result for m will always be something the same, since you have to go through the whole array to even find the minimum.

And I thought if i dont use m, where do i store the the minimum value in the array????

It's not storage that is the real problem, it's just that static variables of any type are very inflexible. A static variable can only be initialized once.

>> if(!m) m = arr;
This will only be true once, in fact, it's only true the first time it's tested. You can't ever use min on a different array.

The trick with recursion is breaking the problem to the first step and letting the function do the rest.

Define the base case, in this case maybe once you reach the end or the array.

Solve for the very first case, as in testing array element 0.

Maybe something like this...

Code:

int min(int *arr,int size) //might want to consider in passing in three args (int *arr, int size, int accumulator)
{
   //Accumulator (acc)
   //set the acc equal to the first element of the array
   //if at the array end(assuming size in your case) return the acc.
  
  //call the function again, comparing the acc to the next element of the array.
}

@grumpy:

That min() should never be called with size zero. It eventually will be.

Sorry I could not actually catch what or why is it for?! Could you please take the trouble of explaining it to me again.

@whiteflags: The program was run on -
1. Pelles C x64 on Windows 7 Ultimate SP1 (64 bit) - it always gave the same error for different sets of inputs, It always prints, The minimum no. is 0
2. Dev-CPP (32 bit) on same platform - it gave correct answers for all different sets of input.

I cant find any logical or technical errors in my code. So what's actually going wrong.....???????????

@all: thanks and would appreciate further help on this....if any one has a definite reason for this nonsense....

Originally Posted by Avenger625

1. Pelles C x64 on Windows 7 Ultimate SP1 (64 bit) - it always gave the same error for different sets of inputs, It always prints, The minimum no. is 0
2. Dev-CPP (32 bit) on same platform - it gave correct answers for all different sets of input.

This is just a guess, but it might be how things are being computed via 64/32 bit.

One option you can try is, using the Debug mode of Pelles (assuming it has one) and step through your code line by line. And see how values are changing and what they are changing to.

@grumpy: I see. Thanks. It works fine now when min() is not called with size 0. But my question is - Why did Dev-Cpp work fine with that but not Pelles C x64?????