Probably by using IEEE 754. What will make any function that uses floating point numbers mindblowing complex. But there's probably an easier way by creating so

Probably by using IEEE 754. What will make any function that uses floating point numbers mindblowing complex. But there's probably an easier way by creating some transformation from (Integer -> a) to (F64 -> a) so that only the transformation gets complex. Anyway, there are many reasons people don't write actual programs this way.