Portability GHC only experimental ekmett@gmail.com None

Contents

Description

Reverse-mode automatic differentiation using Wengert lists and Data.Reflection

Synopsis

Documentation

data Reverse a s Source

Instances

 Typeable2 Reverse (Reifies * s Tape, Num a, Bounded a) => Bounded (Reverse a s) (Reifies * s Tape, Num a, Enum a) => Enum (Reverse a s) (Reifies * s Tape, Num a, Eq a) => Eq (Reverse a s) (Reifies * s Tape, Floating a) => Floating (Reverse a s) (Reifies * s Tape, Fractional a) => Fractional (Reverse a s) (Reifies * s Tape, Num a) => Num (Reverse a s) (Reifies * s Tape, Num a, Ord a) => Ord (Reverse a s) (Reifies * s Tape, Real a) => Real (Reverse a s) (Reifies * s Tape, RealFloat a) => RealFloat (Reverse a s) (Reifies * s Tape, RealFrac a) => RealFrac (Reverse a s) Show a => Show (Reverse a s) (Reifies * s Tape, Erf a) => Erf (Reverse a s) (Reifies * s Tape, InvErf a) => InvErf (Reverse a s) (Num a, Reifies * s Tape) => Mode (Reverse a s) (Reifies * s Tape, Num a) => Jacobian (Reverse a s)

grad :: (Traversable f, Num a) => (forall s. Reifies s Tape => f (Reverse a s) -> Reverse a s) -> f a -> f aSource

The grad function calculates the gradient of a non-scalar-to-scalar function with reverse-mode AD in a single pass.

>>> grad (\[x,y,z] -> x*y+z) [1,2,3]
[2,1,1]

grad' :: (Traversable f, Num a) => (forall s. Reifies s Tape => f (Reverse a s) -> Reverse a s) -> f a -> (a, f a)Source

The grad' function calculates the result and gradient of a non-scalar-to-scalar function with reverse-mode AD in a single pass.

>>> grad' (\[x,y,z] -> x*y+z) [1,2,3]
(5,[2,1,1])

gradWith :: (Traversable f, Num a) => (a -> a -> b) -> (forall s. Reifies s Tape => f (Reverse a s) -> Reverse a s) -> f a -> f bSource

grad g f function calculates the gradient of a non-scalar-to-scalar function f with reverse-mode AD in a single pass. The gradient is combined element-wise with the argument using the function g.

gradWith' :: (Traversable f, Num a) => (a -> a -> b) -> (forall s. Reifies s Tape => f (Reverse a s) -> Reverse a s) -> f a -> (a, f b)Source

grad' g f calculates the result and gradient of a non-scalar-to-scalar function f with reverse-mode AD in a single pass the gradient is combined element-wise with the argument using the function g.

Jacobian

jacobian :: (Traversable f, Functor g, Num a) => (forall s. Reifies s Tape => f (Reverse a s) -> g (Reverse a s)) -> f a -> g (f a)Source

The jacobian function calculates the jacobian of a non-scalar-to-non-scalar function with reverse AD lazily in m passes for m outputs.

>>> jacobian (\[x,y] -> [y,x,x*y]) [2,1]
[[0,1],[1,0],[1,2]]

jacobian' :: (Traversable f, Functor g, Num a) => (forall s. Reifies s Tape => f (Reverse a s) -> g (Reverse a s)) -> f a -> g (a, f a)Source

The jacobian' function calculates both the result and the Jacobian of a nonscalar-to-nonscalar function, using m invocations of reverse AD, where m is the output dimensionality. Applying fmap snd to the result will recover the result of jacobian | An alias for gradF'

>>> jacobian' (\[x,y] -> [y,x,x*y]) [2,1]
[(1,[0,1]),(2,[1,0]),(2,[1,2])]

jacobianWith :: (Traversable f, Functor g, Num a) => (a -> a -> b) -> (forall s. Reifies s Tape => f (Reverse a s) -> g (Reverse a s)) -> f a -> g (f b)Source

'jacobianWith g f' calculates the Jacobian of a non-scalar-to-non-scalar function f with reverse AD lazily in m passes for m outputs.

Instead of returning the Jacobian matrix, the elements of the matrix are combined with the input using the g.

jacobian == jacobianWith (_ dx -> dx)
jacobianWith const == (f x -> const x <\$> f x)

jacobianWith' :: (Traversable f, Functor g, Num a) => (a -> a -> b) -> (forall s. Reifies s Tape => f (Reverse a s) -> g (Reverse a s)) -> f a -> g (a, f b)Source

jacobianWith g f' calculates both the result and the Jacobian of a nonscalar-to-nonscalar function f, using m invocations of reverse AD, where m is the output dimensionality. Applying fmap snd to the result will recover the result of jacobianWith

Instead of returning the Jacobian matrix, the elements of the matrix are combined with the input using the g.

jacobian' == jacobianWith' (_ dx -> dx)

Hessian

hessian :: (Traversable f, Num a) => (forall s s'. (Reifies s Tape, Reifies s' Tape) => f (On (Reverse (Reverse a s') s)) -> On (Reverse (Reverse a s') s)) -> f a -> f (f a)Source

Compute the hessian via the jacobian of the gradient. gradient is computed in reverse mode and then the jacobian is computed in reverse mode.

However, since the grad f :: f a -> f a is square this is not as fast as using the forward-mode Jacobian of a reverse mode gradient provided by hessian.

>>> hessian (\[x,y] -> x*y) [1,2]
[[0,1],[1,0]]

hessianF :: (Traversable f, Functor g, Num a) => (forall s s'. (Reifies s Tape, Reifies s' Tape) => f (On (Reverse (Reverse a s') s)) -> g (On (Reverse (Reverse a s') s))) -> f a -> g (f (f a))Source

Compute the order 3 Hessian tensor on a non-scalar-to-non-scalar function via the reverse-mode Jacobian of the reverse-mode Jacobian of the function.

Less efficient than hessianF.

>>> hessianF (\[x,y] -> [x*y,x+y,exp x*cos y]) [1,2]
[[[0.0,1.0],[1.0,0.0]],[[0.0,0.0],[0.0,0.0]],[[-1.1312043837568135,-2.4717266720048188],[-2.4717266720048188,1.1312043837568135]]]

Derivatives

diff :: Num a => (forall s. Reifies s Tape => Reverse a s -> Reverse a s) -> a -> aSource

Compute the derivative of a function.

>>> diff sin 0
1.0

diff' :: Num a => (forall s. Reifies s Tape => Reverse a s -> Reverse a s) -> a -> (a, a)Source

The diff' function calculates the result and derivative, as a pair, of a scalar-to-scalar function.

>>> diff' sin 0
(0.0,1.0)
>>> diff' exp 0
(1.0,1.0)

diffF :: (Functor f, Num a) => (forall s. Reifies s Tape => Reverse a s -> f (Reverse a s)) -> a -> f aSource

Compute the derivatives of each result of a scalar-to-vector function with regards to its input.

>>> diffF (\a -> [sin a, cos a]) 0
[1.0,0.0]

diffF' :: (Functor f, Num a) => (forall s. Reifies s Tape => Reverse a s -> f (Reverse a s)) -> a -> f (a, a)Source

Compute the derivatives of each result of a scalar-to-vector function with regards to its input along with the answer.

>>> diffF' (\a -> [sin a, cos a]) 0
[(0.0,1.0),(1.0,0.0)]