| Copyright | (C) 2014 Edward Kmett |
|---|---|
| License | BSD-style (see the file LICENSE) |
| Maintainer | Edward Kmett <ekmett@gmail.com> |
| Stability | experimental |
| Portability | PatternSynonyms |
| Safe Haskell | Trustworthy |
| Language | Haskell2010 |
Numeric.Half.Internal
Contents
Description
Half-precision floating-point values. These arise commonly in GPU work and it is useful to be able to compute them and compute with them on the CPU as well.
Synopsis
- newtype Half = Half {}
- isZero :: Half -> Bool
- fromHalf :: Half -> Float
- toHalf :: Float -> Half
- pattern POS_INF :: Half
- pattern NEG_INF :: Half
- pattern QNaN :: Half
- pattern SNaN :: Half
- pattern HALF_MIN :: Half
- pattern HALF_NRM_MIN :: Half
- pattern HALF_MAX :: Half
- pattern HALF_EPSILON :: Half
- pattern HALF_DIG :: (Eq a, Num a) => a
- pattern HALF_MIN_10_EXP :: (Eq a, Num a) => a
- pattern HALF_MAX_10_EXP :: (Eq a, Num a) => a
- pure_floatToHalf :: Float -> Half
- pure_halfToFloat :: Half -> Float
Documentation
Instances
Patterns
These are available with GHC-7.8 and later.
pattern HALF_NRM_MIN :: Half Source #
Smallest positive normalized half
pattern HALF_EPSILON :: Half Source #
Smallest positive e for which half (1.0 + e) != half (1.0)
pattern HALF_DIG :: (Eq a, Num a) => a Source #
Number of base 10 digits that can be represented without change
pattern HALF_MIN_10_EXP :: (Eq a, Num a) => a Source #
pattern HALF_MAX_10_EXP :: (Eq a, Num a) => a Source #