Previous: Rule of Float Underflow and Overflow, Up: Floating-point Computations
The result of a numerical function is a float of the largest format among all the floating-point arguments to the function.