Why does pow(x,y) return inf when x and y are chars?

Question

I understand how to use pow() correctly I'm just wondering why when I run this code ans = inf.

I am having a hard time understanding this.

Does this have anything to do with chars only being able to take on the values -128 to +127 and the way pow() is calculated?

Does this have anything to do with the space in " %c" as in my first scanf param?

Linux 4.9.0-7-amd64 debian 4.9.110-1 gcc version 6.3.020170516

#include <stdio.h>
#include <math.h>

int main ()
{

    char base, exp; 
    float ans; 

    printf("Enter base number : ");
    scanf(" %c", &base); 

    printf("Enter exp number : ");
    scanf(" %c", &exp); 

    ans = pow(base,exp); 

    printf("%c raised to the %c power equals %f", base, exp, ans);

    return 0; 
}

Answer 1

When you enter base and exp using %c you get the characters '0' - '9', not the integer values. So if you enter 0 and 0, the values you get (assuming ASCII encoding) will actually by 48, since that is the ASCII code for '0'. The result of 48⁴⁸ is roughly 5 x 10⁸⁰.

You then save the resulting value in a float. Assuming a float is IEEE754 single precision, it can hold an exponent no larger that +/- 38. So you're attempting convert an out-of-range value which invokes undefined behavior, meaning the result is unpredictable. If you changed the type of ans to double you would see the actual result.

Answer 2

Base and exp is not the integer values (0-9). So you want to enter "%d". Not the " %c"