Recursive algorithim...returns a 0, whats wrong with the formula???

[hotdogs113]

I'm writing a 3 function program for homework. The first and third functions seem to work fine, but the second does not.

The function is a recursive function that should do x to n power. It works for whatever x is, and only up to the power of 5. After that, it always returns a 0.

Here's the code...any advice would be appreciated.

Function in questions starts at line 18.

# include <iostream>


using namespace std;



//void loop (int x,int n)
//{

	//int Bosco=1;
	//for (int i=1;i<n;i++)
	//	Bosco=Bosco*x;
//cout<<Bosco<<endl;

//}

int rec(int x,int n)
{
	
	if (n==0)
	{return (1);}
	else if (n==1){ return (x);}
	else if (n)
	{return rec ((x*x),n-1);}
	}

float power (int x,int n)
{
	if (n==0){return 1;}
	else if (n==1) {return x;}
	else if (n)
		return power (x*x,n/2);
	else return power (x*x,n/2)*x;
}

int main ()

{ 
	int x;
	int n;
	int a,b;
	
	cout<<"Enter a number"<<endl;
cin>>x;
cout<<"Enter a number"<<endl;
cin>>n;

//loop (x,n);
rec(x,n);
power (x,n);
a=rec (x,n);
b=power (x,n);
cout<<a<<"and"<<b<<endl;

	cout<<"Thanks"<<endl;
}

[hamsterman]

Only your first function is good.
What the second one does is not xⁿ, but x^{2^n} (^ is exponentiation. Apparently you can't have nested sup tags here..), so when n=5, 2⁵=32 and for any x!=1, x³² is more than fits in an integer.
The correct way is return rec(x, n-1)*x;
I assume the purpose of the third one is a faster algorithm where x⁴ is (x²)², and this should work fine. However, what if you have x⁶? Then it should be ((x²)*x))². I suppose that's the purpose of line 34, but it will never be reached, since (n) evaluates to true when n != 0. Did you mean (n%2 == 0)?

[hotdogs113]

Hamsterman....odd question here...I put in your recommended change and it worked. I then took the // away and ran the full three functions. Function 1 and 2 were supposed to give the same output.

I then went into function 1 and changed the int i = 1 to i=0, and now they both give the same output. I then double checked and they are giving the correct answer.

Thanks!

The third function....I took it verbatim out of my class notes the professor provided.....and when I put in the test numbers the output seems correct.

When I looked at it again, you are right, the (n) was wrong, it needed the n%==2.

Thanks again...

[hotdogs113]

Well interesting development....

When I ran the program with the given input, the last was x=2, n=50.

It goes nuts.......

[hamsterman]

integer is 32 bits. 1 for sign (sort of), 31 for the number, so the maximum value you can calculate with these functions is 2³¹. If you feel that you need more, see http://gmplib.org/

[hotdogs113]

So do you think my professor is being tricky here?

[hamsterman]

You could use long long int to increase the limit to 2⁶⁴..

[hotdogs113]

Hm tried that and some other changes, still coming off as 0....

[hamsterman]

x also has to be long long int. That's because C++ operators preserve type. int*int will always return int, even if you assign it to long long int. For operator * to return long long int, at least one of it's arguments has to be one too.

[hotdogs113]

It works! Thanks alot!