Function dfactorial

libraries/statHelpers/statHelpers.cpp:165–177

FASTER VERSION does part of the math with integers. tested on UNO and ESP32, roughly 3x faster numbers differ slightly in the order of IEEE754 precision => acceptable. 10e-7 for 4 bit float 10e-16 for 8 bit double

//  10e-7  for 4 bit float
//  10e-16 for 8 bit double
double dfactorial(uint8_t n)
{
  double f = 1;
  while (n > 4)
  {
    uint32_t val = n * (n-1);
    val *= (n-2) * (n-3);
    f *= val;
    n -= 4;
  }
  while (n > 1) f *= (n--);   //  can be squeezed too.
  return f;
}


//  stirling is an approximation function for factorial(n).