The number of bits used to represent the type’s exponent.
- Xcode 8.0+
- Swift Standard Library
A binary floating-point type’s
exponent imposes a limit on the range of the exponent for normal, finite values. The exponent bias of a type
F can be calculated as the following, where
** is exponentiation:
The least normal exponent for values of the type
1 - bias, and the largest finite exponent is
bias. An all-zeros exponent is reserved for subnormals and zeros, and an all-ones exponent is reserved for infinity and NaN.
For example, the
Float type has an
exponent of 8, which gives an exponent bias of
127 by the calculation above.