Evaluate a rational function using single-precision floating-point arithmetic. - View it on GitHub