Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
branch: master
Fetching contributors…

Cannot retrieve contributors at this time

file 162 lines (148 sloc) 5.076 kb
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162
name: ad
version: 1.1.2
license: BSD3
license-File: LICENSE
copyright: (c) Edward Kmett 2010-2011,
              (c) Barak Pearlmutter and Jeffrey Mark Siskind 2008-2009
author: Edward Kmett
maintainer: ekmett@gmail.com
stability: Experimental
category: Math
homepage: http://github.com/ekmett/ad
build-type: Simple
cabal-version: >= 1.6
extra-source-files: TODO
synopsis: Automatic Differentiation

description:
    Forward-, reverse- and mixed- mode automatic differentiation combinators with a common API.
    .
    Type-level \"branding\" is used to both prevent the end user from confusing infinitesimals
    and to limit unsafe access to the implementation details of each Mode.
    .
    Each mode has a separate module full of combinators.
    .
    * @Numeric.AD.Mode.Forward@ provides basic forward-mode AD. It is good for computing simple derivatives.
    .
    * @Numeric.AD.Mode.Reverse@ uses benign side-effects to compute reverse-mode AD. It is good for computing gradients in one pass.
    .
    * @Numeric.AD.Mode.Sparse@ computes a sparse forward-mode AD tower. It is good for higher derivatives or large numbers of outputs.
    .
    * @Numeric.AD.Mode.Tower@ computes a dense forward-mode AD tower useful for higher derivatives of single input functions.
    .
    * @Numeric.AD.Mode.Mixed@ computes using whichever mode or combination thereof is suitable to each individual combinator. This mode is the default, re-exported by @Numeric.AD@
    .
    .
    While not every mode can provide all operations, the following basic operations are supported, modified as
    appropriate by the suffixes below:
    .
    * 'grad' computes the gradient (partial derivatives) of a function at a point.
    .
    * 'jacobian' computes the Jacobian matrix of a function at a point.
    .
    * 'diff' computes the derivative of a function at a point.
    .
    * 'du' computes a directional derivative of a function at a point.
    .
    * 'hessian' computes the Hessian matrix (matrix of second partial derivatives) of a function at a point.
    .
    The following suffixes alter the meanings of the functions above as follows:
    .
    * @\'@ -- also return the answer
    .
    * @With@ lets the user supply a function to blend the input with the output
    .
    * @F@ is a version of the base function lifted to return a 'Traversable' (or 'Functor') result
    .
    * @s@ means the function returns all higher derivatives in a list or f-branching 'Stream'
    .
    * @T@ means the result is transposed with respect to the traditional formulation.
    .
    * @0@ means that the resulting derivative list is padded with 0s at the end.
    .
    Changes since 1.0.0
    .
    * Changed the way 'Show' was derived to comply with changes in instance resolution in ghc >= 7.0 && <= 7.1
    .
    Changes since 0.45.0
    .
    * Converted 'Stream' to use the external 'comonad' package
    .
    Changes since 0.44.5
    .
    * Added Halley's method
    .
    Changes since 0.40.0
    .
    * Fixed bug fix for @'(/)' :: (Mode s, Fractional a) => AD s a@
    .
    * Improved documentation
    .
    * Regularized naming conventions
    .
    * Exposed 'Id', probe, and lower methods via @Numeric.AD.Types@
    .
    * Removed monadic combinators
    .
    * Retuned the 'Mixed' mode jacobian calculations to only require a 'Functor'-based result.
    .
    * Added unsafe variadic 'vgrad', 'vgrad'', and 'vgrads' combinators


source-repository head
  type: git
  location: git://github.com/ekmett/ad.git

flag TemplateHaskell24
  manual: False
  default: False

library
  extensions: CPP

  other-extensions:
    BangPatterns
    DeriveDataTypeable
    FlexibleContexts
    FlexibleInstances
    FunctionalDependencies
    GeneralizedNewtypeDeriving
    MultiParamTypeClasses
    PatternGuards
    Rank2Types
    ScopedTypeVariables
    TemplateHaskell
    TypeFamilies
    TypeOperators
    UndecidableInstances

  if flag(TemplateHaskell24)
    build-depends: template-haskell >= 2.4 && < 2.5
    cpp-options: -DOldClassI
  else
    build-depends: template-haskell >= 2.5 && < 2.6
  
  build-depends:
    base >= 4 && < 5,
    data-reify >= 0.6 && < 0.7,
    containers >= 0.2 && < 0.5,
    array >= 0.2 && < 0.4,
    comonad >= 1.1 && < 1.2,
    free >= 1.8 && < 1.9
  
  exposed-modules:
    Numeric.AD
    Numeric.AD.Classes
    Numeric.AD.Types
    Numeric.AD.Newton
    Numeric.AD.Halley
  
    Numeric.AD.Internal.Classes
    Numeric.AD.Internal.Combinators
  
    Numeric.AD.Internal.Forward
    Numeric.AD.Internal.Tower
    Numeric.AD.Internal.Reverse
    Numeric.AD.Internal.Sparse
    Numeric.AD.Internal.Dense
    Numeric.AD.Internal.Composition
  
    Numeric.AD.Mode.Directed
    Numeric.AD.Mode.Forward
    Numeric.AD.Mode.Mixed
    Numeric.AD.Mode.Reverse
    Numeric.AD.Mode.Tower
    Numeric.AD.Mode.Sparse
  
  other-modules:
    Numeric.AD.Internal.Types
    Numeric.AD.Internal.Tensors
    Numeric.AD.Internal.Identity
  
  ghc-options: -Wall -fspec-constr -fdicts-cheap -O2
Something went wrong with that request. Please try again.