Fractional Exponents are Non-Calculable about graphing-calculator-project HOT 3 OPEN

The main reason for the truncation was because of a highlighted part of this code.
The built in library that handles all the Big-Numbers is truncating the whole exponent to a 32 bit integer.
Which isn't necessarily bad, but kind of absurd if you ask me because most, if not all floating point decimals are gone.

Revised log tables will be needed for all of this math to work. The Epsilon will have to be extended.
long floats are required. (in concept, but that wont actually be the case)

void
bc_raise (num1, num2, result, scale)
bc_num num1, num2, *result;
int scale;
{
bc_num temp, power;
long exponent;
int rscale;
int pwrscale;
int calcscale;
char neg;

_/ Check the exponent for scale digits and convert to a long. /
if (num2->n_scale != 0)
bc_rt_warn (BC_WARNING_NON_ZERO_SCALE_IN_EXPONENT);
exponent = bc_num2long (num2);
if (exponent == 0 && (num2->n_len > 1 || num2->n_value[0] != 0))
bc_rt_error (BC_ERROR_EXPONENT_TOO_LARGE_IN_RAISE);_

/* Special case if exponent is a zero. */
if (exponent == 0)
{
bc_free_num (result);
*result = bc_copy_num (one);
return;
}

/* Other initializations. /
if (exponent < 0)
{
neg = TRUE;
exponent = -exponent;
rscale = scale;
}
else
{
neg = FALSE;
rscale = MIN (num1->n_scaleexponent, MAX(scale, num1->n_scale));
}

/* Set initial value of temp. /
power = bc_copy_num (num1);
pwrscale = num1->n_scale;
while ((exponent & 1) == 0)
{
pwrscale = 2pwrscale;
bc_multiply (power, power, &power, pwrscale);
exponent = exponent >> 1;
}
temp = bc_copy_num (power);
calcscale = pwrscale;
exponent = exponent >> 1;

/* Do the calculation. /
while (exponent > 0)
{
pwrscale = 2pwrscale;
bc_multiply (power, power, &power, pwrscale);
if ((exponent & 1) == 1) {
calcscale = pwrscale + calcscale;
bc_multiply (temp, power, &temp, calcscale);
}
exponent = exponent >> 1;
}

/* Assign the value. */
if (neg)
{
bc_divide (one, temp, result, rscale);
bc_free_num (&temp);
}
else
{
bc_free_num (result);
*result = temp;
if ((*result)->n_scale > rscale)
(*result)->n_scale = rscale;
}
bc_free_num (&power);
}

from graphing-calculator-project.

https://www.arduino.cc/en/Math/H
Further insight: Look into how doubles are handled and update the built in log table for arduino. (If I have to)
double atan (double __x) // arc tangent of x
double atan2 (double __y, double __x) // arc tangent of y/x
double log (double __x) // natural logarithm of x
double log10 (double __x) // logarithm of x to base 10.
double pow (double __x, double __y) // x to power of y
double square (double __x) // square of x

from graphing-calculator-project.

http://stackoverflow.com/questions/34239178/how-does-an-avr-perform-floating-point-arithmetic
It would be good to learn from the resources in this thread.

from graphing-calculator-project.