Compute the relative difference of two real numbers in units of double-precision floating-point epsilon. - View it on GitHub