Copyright | (c) Edward Kmett 2010-2021 |
---|---|
License | BSD3 |
Maintainer | ekmett@gmail.com |
Stability | experimental |
Portability | GHC only |
Safe Haskell | Safe-Inferred |
Language | Haskell2010 |
A version of Double
that rounds to the twelfth digit if necessary. This
is useful for ad
's doctests since they must print out floating-point
numbers in their entirety, but the actual numbers that get produced can
vary slightly depending on machine-specific implementation details.
(See #73 for an example.) This works around the issue by just rounding
up the printed result to a point where it should be consistent across
all machines.