The most general data type for a rational number stores the numerator and denominator as integers. See integral data type.
A "fixed point" data type assumes a specific denominator for all numbers. The denominator here is most often a power of two. For example, in a system whose denominator is 65,536 (2^{16}), the hexadecimal number 0x12345678 means 0x12345678/65536 or 305419896/65536 or 4660 + 22136/65536 or about 4660.33777. See fixed point.
A "floating point" type is a compromise between the flexibility of a general rational type and the speed of fixedpoint arithmetic. It uses some of the bits in the data type to specify a power of two for the denominator. See floating point and IEEE Floating Point Standard.
Search Encyclopedia

Featured Article
