I feel like a lot of conditionals involving floats make more sense if you convert it to an integer and do bitwise operations lmao. Like to check if something is a real number just check to see if the exponent is not in the space used by infinities and nans