math-functions-0.3.4.2: Collection of tools for numeric computations
Copyright (c) 2014 Bryan O'Sullivan
License BSD3
Maintainer bos@serpentine.com
Stability experimental
Portability portable
Safe Haskell None
Language Haskell2010

Numeric.Sum

Description

Functions for summing floating point numbers more accurately than the naive sum function and its counterparts in the vector package and elsewhere.

When used with floating point numbers, in the worst case, the sum function accumulates numeric error at a rate proportional to the number of values being summed. The algorithms in this module implement different methods of /compensated summation/, which reduce the accumulation of numeric error so that it either grows much more slowly than the number of inputs (e.g. logarithmically), or remains constant.

Synopsis

Summation type class

sumVector :: ( Vector v Double , Summation s) => (s -> Double ) -> v Double -> Double Source #

O(n) Sum a vector of values.

Usage

Most of these summation algorithms are intended to be used via the Summation typeclass interface. Explicit type annotations should not be necessary, as the use of a function such as kbn or kb2 to extract the final sum out of a Summation instance gives the compiler enough information to determine the precise type of summation algorithm to use.

As an example, here is a (somewhat silly) function that manually computes the sum of elements in a list.

sillySumList :: [Double] -> Double
sillySumList = loop zero
  where loop s []     = kbn s
        loop s (x:xs) = seq s' loop s' xs
          where s'    = add s x

In most instances, you can simply use the much more general sum function instead of writing a summation function by hand.

-- Avoid ambiguity around which sum function we are using.
import Prelude hiding (sum)
--
betterSumList :: [Double] -> Double
betterSumList xs = sum kbn xs

Kahan-Babuška-Neumaier summation

data KBNSum Source #

Kahan-Babuška-Neumaier summation. This is a little more computationally costly than plain Kahan summation, but is always at least as accurate.

Instances

Instances details
Eq KBNSum Source #
Instance details

Defined in Numeric.Sum

Data KBNSum Source #
Instance details

Defined in Numeric.Sum

Methods

gfoldl :: ( forall d b. Data d => c (d -> b) -> d -> c b) -> ( forall g. g -> c g) -> KBNSum -> c KBNSum Source #

gunfold :: ( forall b r. Data b => c (b -> r) -> c r) -> ( forall r. r -> c r) -> Constr -> c KBNSum Source #

toConstr :: KBNSum -> Constr Source #

dataTypeOf :: KBNSum -> DataType Source #

dataCast1 :: Typeable t => ( forall d. Data d => c (t d)) -> Maybe (c KBNSum ) Source #

dataCast2 :: Typeable t => ( forall d e. ( Data d, Data e) => c (t d e)) -> Maybe (c KBNSum ) Source #

gmapT :: ( forall b. Data b => b -> b) -> KBNSum -> KBNSum Source #

gmapQl :: (r -> r' -> r) -> r -> ( forall d. Data d => d -> r') -> KBNSum -> r Source #

gmapQr :: forall r r'. (r' -> r -> r) -> r -> ( forall d. Data d => d -> r') -> KBNSum -> r Source #

gmapQ :: ( forall d. Data d => d -> u) -> KBNSum -> [u] Source #

gmapQi :: Int -> ( forall d. Data d => d -> u) -> KBNSum -> u Source #

gmapM :: Monad m => ( forall d. Data d => d -> m d) -> KBNSum -> m KBNSum Source #

gmapMp :: MonadPlus m => ( forall d. Data d => d -> m d) -> KBNSum -> m KBNSum Source #

gmapMo :: MonadPlus m => ( forall d. Data d => d -> m d) -> KBNSum -> m KBNSum Source #

Show KBNSum Source #
Instance details

Defined in Numeric.Sum

Semigroup KBNSum Source #

Since: 0.3.0.0

Instance details

Defined in Numeric.Sum

Monoid KBNSum Source #

Since: 0.3.0.0

Instance details

Defined in Numeric.Sum

NFData KBNSum Source #
Instance details

Defined in Numeric.Sum

Unbox KBNSum Source #
Instance details

Defined in Numeric.Sum

Summation KBNSum Source #
Instance details

Defined in Numeric.Sum

Vector Vector KBNSum Source #
Instance details

Defined in Numeric.Sum

MVector MVector KBNSum Source #
Instance details

Defined in Numeric.Sum

newtype Vector KBNSum Source #
Instance details

Defined in Numeric.Sum

newtype MVector s KBNSum Source #
Instance details

Defined in Numeric.Sum

kbn :: KBNSum -> Double Source #

Return the result of a Kahan-Babuška-Neumaier sum.

Order-2 Kahan-Babuška summation

data KB2Sum Source #

Second-order Kahan-Babuška summation. This is more computationally costly than Kahan-Babuška-Neumaier summation, running at about a third the speed. Its advantage is that it can lose less precision (in admittedly obscure cases).

This method compensates for error in both the sum and the first-order compensation term, hence the use of "second order" in the name.

Instances

Instances details
Eq KB2Sum Source #
Instance details

Defined in Numeric.Sum

Data KB2Sum Source #
Instance details

Defined in Numeric.Sum

Methods

gfoldl :: ( forall d b. Data d => c (d -> b) -> d -> c b) -> ( forall g. g -> c g) -> KB2Sum -> c KB2Sum Source #

gunfold :: ( forall b r. Data b => c (b -> r) -> c r) -> ( forall r. r -> c r) -> Constr -> c KB2Sum Source #

toConstr :: KB2Sum -> Constr Source #

dataTypeOf :: KB2Sum -> DataType Source #

dataCast1 :: Typeable t => ( forall d. Data d => c (t d)) -> Maybe (c KB2Sum ) Source #

dataCast2 :: Typeable t => ( forall d e. ( Data d, Data e) => c (t d e)) -> Maybe (c KB2Sum ) Source #

gmapT :: ( forall b. Data b => b -> b) -> KB2Sum -> KB2Sum Source #

gmapQl :: (r -> r' -> r) -> r -> ( forall d. Data d => d -> r') -> KB2Sum -> r Source #

gmapQr :: forall r r'. (r' -> r -> r) -> r -> ( forall d. Data d => d -> r') -> KB2Sum -> r Source #

gmapQ :: ( forall d. Data d => d -> u) -> KB2Sum -> [u] Source #

gmapQi :: Int -> ( forall d. Data d => d -> u) -> KB2Sum -> u Source #

gmapM :: Monad m => ( forall d. Data d => d -> m d) -> KB2Sum -> m KB2Sum Source #

gmapMp :: MonadPlus m => ( forall d. Data d => d -> m d) -> KB2Sum -> m KB2Sum Source #

gmapMo :: MonadPlus m => ( forall d. Data d => d -> m d) -> KB2Sum -> m KB2Sum Source #

Show KB2Sum Source #
Instance details

Defined in Numeric.Sum

Semigroup KB2Sum Source #

Since: 0.3.0.0

Instance details

Defined in Numeric.Sum

Monoid KB2Sum Source #

Since: 0.3.0.0

Instance details

Defined in Numeric.Sum

NFData KB2Sum Source #
Instance details

Defined in Numeric.Sum

Unbox KB2Sum Source #
Instance details

Defined in Numeric.Sum

Summation KB2Sum Source #
Instance details

Defined in Numeric.Sum

Vector Vector KB2Sum Source #
Instance details

Defined in Numeric.Sum

MVector MVector KB2Sum Source #
Instance details

Defined in Numeric.Sum

newtype Vector KB2Sum Source #
Instance details

Defined in Numeric.Sum

newtype MVector s KB2Sum Source #
Instance details

Defined in Numeric.Sum

kb2 :: KB2Sum -> Double Source #

Return the result of an order-2 Kahan-Babuška sum.

Less desirable approaches

Kahan summation

data KahanSum Source #

Kahan summation. This is the least accurate of the compensated summation methods. In practice, it only beats naive summation for inputs with large magnitude. Kahan summation can be less accurate than naive summation for small-magnitude inputs.

This summation method is included for completeness. Its use is not recommended. In practice, KBNSum is both 30% faster and more accurate.

Instances

Instances details
Eq KahanSum Source #
Instance details

Defined in Numeric.Sum

Data KahanSum Source #
Instance details

Defined in Numeric.Sum

Methods

gfoldl :: ( forall d b. Data d => c (d -> b) -> d -> c b) -> ( forall g. g -> c g) -> KahanSum -> c KahanSum Source #

gunfold :: ( forall b r. Data b => c (b -> r) -> c r) -> ( forall r. r -> c r) -> Constr -> c KahanSum Source #

toConstr :: KahanSum -> Constr Source #

dataTypeOf :: KahanSum -> DataType Source #

dataCast1 :: Typeable t => ( forall d. Data d => c (t d)) -> Maybe (c KahanSum ) Source #

dataCast2 :: Typeable t => ( forall d e. ( Data d, Data e) => c (t d e)) -> Maybe (c KahanSum ) Source #

gmapT :: ( forall b. Data b => b -> b) -> KahanSum -> KahanSum Source #

gmapQl :: (r -> r' -> r) -> r -> ( forall d. Data d => d -> r') -> KahanSum -> r Source #

gmapQr :: forall r r'. (r' -> r -> r) -> r -> ( forall d. Data d => d -> r') -> KahanSum -> r Source #

gmapQ :: ( forall d. Data d => d -> u) -> KahanSum -> [u] Source #

gmapQi :: Int -> ( forall d. Data d => d -> u) -> KahanSum -> u Source #

gmapM :: Monad m => ( forall d. Data d => d -> m d) -> KahanSum -> m KahanSum Source #

gmapMp :: MonadPlus m => ( forall d. Data d => d -> m d) -> KahanSum -> m KahanSum Source #

gmapMo :: MonadPlus m => ( forall d. Data d => d -> m d) -> KahanSum -> m KahanSum Source #

Show KahanSum Source #
Instance details

Defined in Numeric.Sum

Semigroup KahanSum Source #

Since: 0.3.0.0

Instance details

Defined in Numeric.Sum

Monoid KahanSum Source #

Since: 0.3.0.0

Instance details

Defined in Numeric.Sum

NFData KahanSum Source #
Instance details

Defined in Numeric.Sum

Unbox KahanSum Source #
Instance details

Defined in Numeric.Sum

Summation KahanSum Source #
Instance details

Defined in Numeric.Sum

Vector Vector KahanSum Source #
Instance details

Defined in Numeric.Sum

MVector MVector KahanSum Source #
Instance details

Defined in Numeric.Sum

newtype Vector KahanSum Source #
Instance details

Defined in Numeric.Sum

newtype MVector s KahanSum Source #
Instance details

Defined in Numeric.Sum

kahan :: KahanSum -> Double Source #

Return the result of a Kahan sum.

Pairwise summation

pairwiseSum :: Vector v Double => v Double -> Double Source #

O(n) Sum a vector of values using pairwise summation.

This approach is perhaps 10% faster than KBNSum , but has poorer bounds on its error growth. Instead of having roughly constant error regardless of the size of the input vector, in the worst case its accumulated error grows with O(log n) .

References

  • Kahan, W. (1965), Further remarks on reducing truncation errors. Communications of the ACM 8(1):40.
  • Neumaier, A. (1974), Rundungsfehleranalyse einiger Verfahren zur Summation endlicher Summen. Zeitschrift für Angewandte Mathematik und Mechanik 54:39–51.
  • Klein, A. (2006), A Generalized Kahan-Babuška-Summation-Algorithm. Computing 76(3):279-293.
  • Higham, N.J. (1993), The accuracy of floating point summation. SIAM Journal on Scientific Computing 14(4):783–799.