Using power function with big numbers

[Gus007]

Hi all,

Need the community great support once more. :)

I need to know how to calculate the power of some numbers in C, the problem is that the number is too big , and the compiler gives a error when doing the arithmetics.

See the power function I am using:

unsigned long int power(int base, int n)
{
unsigned long long int p;
int i;

p = 1;

for (i = 1; i <= n; ++i)
p = p * base;
return p;
}


Passing the parameter 2 as base, I can't execute this function with a n greater than 32... the problem is that I need to calculate from 2**0 to 2**64.

Hope I made myself clear about what I need
Any suggestions?

Thanks!
Gus

[willakawill]

Hi all,

Need the community great support once more. :)

I need to know how to calculate the power of some numbers in C, the problem is that the number is too big , and the compiler gives a error when doing the arithmetics.

See the power function I am using:

unsigned long int power(int base, int n)
{
unsigned long long int p;
int i;

p = 1;

for (i = 1; i <= n; ++i)
p = p * base;
return p;
}


Passing the parameter 2 as base, I can't execute this function with a n greater than 32... the problem is that I need to calculate from 2**0 to 2**64.

Hope I made myself clear about what I need
Any suggestions?

Thanks!
Gus

Hi Gus, the answer is in the question. The built in data type that can handle 2^64 (64 bits) is a double

[Banfa]

Hi Gus, the answer is in the question. The built in data type that can handle 2^64 (64 bits) is a double

Erm many systems today also support a 64 bit integer type as well, for instance

long long - *nix
__int64 - Windows

[willakawill]

Erm many systems today also support a 64 bit integer type as well, for instance

long long - *nix
__int64 - Windows

Interesting. Are these built in data types for every system. i.e. are they portable?

[Banfa]

Interesting. Are these built in data types for every system. i.e. are they portable?

Not really, but then strictly speaking neither is any type.

That is you have no guarentee of the number of bits in a long in any system, however you are guarenteed that a type long does exist.


I suspect that it is possible that as more 64 bit systems emerge the standards will be updated.

Of course what should have happened is that for 64 bit systems

char = 8 bits (actually this is an assumption)
short = 16 or 32 bits
int = 32 or 64 bits
long = 64 bits

This would all be within the constraints of the C standard, unfortuneately the compiler writers were so used to the convention

char = 8 bits (actually this is an assumption)
short = 16
int = 16 or 32 bits
long = 32 bits

that they didn't feel they could alter the number of bits in a long or short even though they are not set by the standard which resulted in them all having to introduce a 64 bit integer extension type.

It's a bit of a mess that needs resolving.

[Gus007]

Thanks for the answers!

As you can see I was using 'long long" in my implementation so this was supposed to work if I was a Unix user,right? Since I am a windows user this is probably the problem.

I didn't know that "long long" was system specific...
Gus