-- Hoogle documentation, generated by Haddock
-- See Hoogle, http://www.haskell.org/hoogle/


-- | A minimal Haskell Neural Network Library
--   
--   hnn provides minimal types and functions to create, train and use feed
--   forward neural networks <a>http://www.haskell.org/haskellwiki/HNN</a>
@package hnn
@version 0.1


-- | Neuron module, defining an artificial neuron type and the basical
--   operations we can do on it
module AI.HNN.Neuron

-- | Our Artificial Neuron type
data Neuron
Neuron :: Double -> UArr Double -> (Double -> Double) -> Neuron
threshold :: Neuron -> Double
weights :: Neuron -> UArr Double
func :: Neuron -> Double -> Double

-- | Creates a Neuron with the given threshold, weights and transfer
--   function
createNeuronU :: Double -> UArr Double -> (Double -> Double) -> Neuron

-- | Equivalent to `createNeuronU t ws heavyside'
createNeuronHeavysideU :: Double -> UArr Double -> Neuron

-- | Equivalent to `createNeuronU t ws sigmoid'
createNeuronSigmoidU :: Double -> UArr Double -> Neuron

-- | Same as createNeuronU, with a list instead of an UArr for the weights
--   (converted to UArr anyway)
createNeuron :: Double -> [Double] -> (Double -> Double) -> Neuron

-- | Same as createNeuronHeavysideU, with a list instead of an UArr for the
--   weights (converted to UArr anyway)
createNeuronHeavyside :: Double -> [Double] -> Neuron

-- | Same as createNeuronSigmoidU, with a list instead of an UArr for the
--   weights (converted to UArr anyway)
createNeuronSigmoid :: Double -> [Double] -> Neuron

-- | The Heavyside function
heavyside :: Double -> Double

-- | The Sigmoid function
sigmoid :: Double -> Double

-- | Computes the output of a given Neuron for given inputs
computeU :: Neuron -> UArr Double -> Double

-- | Computes the output of a given Neuron for given inputs
compute :: Neuron -> [Double] -> Double

-- | Trains a neuron with the given sample, of the form (inputs,
--   wanted_result) and the given learning ratio (alpha)
learnSampleU :: Double -> Neuron -> (UArr Double, Double) -> Neuron
learnSample :: Double -> Neuron -> ([Double], Double) -> Neuron

-- | Trains a neuron with the given samples and the given learning ratio
--   (alpha)
learnSamplesU :: Double -> Neuron -> [(UArr Double, Double)] -> Neuron

-- | Trains a neuron with the given samples and the given learning ratio
--   (alpha)
learnSamples :: Double -> Neuron -> [([Double], Double)] -> Neuron
instance Show Neuron


-- | Layer module, defining functions to work on a neural network layer,
--   which is a list of neurons
module AI.HNN.Layer

-- | Creates a layer compound of n neurons with the Sigmoid transfer
--   function, all having the given threshold and weights.
createSigmoidLayerU :: Int -> Double -> UArr Double -> [Neuron]

-- | Creates a layer compound of n neurons with the Heavyside transfer
--   function, all having the given threshold and weights.
createHeavysideLayerU :: Int -> Double -> UArr Double -> [Neuron]

-- | Creates a layer compound of n neurons with the sigmoid transfer
--   function, all having the given threshold and weights.
createSigmoidLayer :: Int -> Double -> [Double] -> [Neuron]

-- | Creates a layer compound of n neurons with the sigmoid transfer
--   function, all having the given threshold and weights.
createHeavysideLayer :: Int -> Double -> [Double] -> [Neuron]

-- | Computes the outputs of each Neuron of the layer
computeLayerU :: [Neuron] -> UArr Double -> UArr Double

-- | Computes the outputs of each Neuron of the layer
computeLayer :: [Neuron] -> [Double] -> [Double]

-- | Trains each neuron with the given sample and the given learning ratio
learnSampleLayerU :: Double -> [Neuron] -> (UArr Double, UArr Double) -> [Neuron]

-- | Trains each neuron with the given sample and the given learning ratio
learnSampleLayer :: Double -> [Neuron] -> ([Double], [Double]) -> [Neuron]

-- | Trains each neuron with the given samples and the given learning ratio
learnSamplesLayerU :: Double -> [Neuron] -> [(UArr Double, UArr Double)] -> [Neuron]

-- | Trains each neuron with the given samples and the given learning ratio
learnSamplesLayer :: Double -> [Neuron] -> [([Double], [Double])] -> [Neuron]

-- | Returns the quadratic error of a layer for a given sample
quadErrorU :: [Neuron] -> (UArr Double, UArr Double) -> Double

-- | Returns the quadratic error of a layer for a given sample
quadError :: [Neuron] -> ([Double], [Double]) -> Double


-- | Net module, defining functions to work on a neural network, which is a
--   list of list of neurons
module AI.HNN.Net
check :: [[Neuron]] -> Bool
nn :: [[Neuron]] -> [[Neuron]]

-- | Computes the output of the given neural net on the given inputs
computeNetU :: [[Neuron]] -> UArr Double -> UArr Double

-- | Computes the output of the given neural net on the given inputs
computeNet :: [[Neuron]] -> [Double] -> [Double]

-- | Returns the quadratic error of the neural network on the given sample
quadErrorNetU :: [[Neuron]] -> (UArr Double, UArr Double) -> Double

-- | Returns the quadratic error of the neural network on the given sample
quadErrorNet :: [[Neuron]] -> ([Double], [Double]) -> Double

-- | Returns the quadratic error of the neural network on the given samples
globalQuadErrorNetU :: [[Neuron]] -> [(UArr Double, UArr Double)] -> Double

-- | Returns the quadratic error of the neural network on the given samples
globalQuadErrorNet :: [[Neuron]] -> [([Double], [Double])] -> Double

-- | Train the given neural network using the backpropagation algorithm on
--   the given sample with the given learning ratio (alpha)
backPropU :: Double -> [[Neuron]] -> (UArr Double, UArr Double) -> [[Neuron]]

-- | Train the given neural network using the backpropagation algorithm on
--   the given sample with the given learning ratio (alpha)
backProp :: Double -> [[Neuron]] -> ([Double], [Double]) -> [[Neuron]]
trainAux :: Double -> [[Neuron]] -> [(UArr Double, UArr Double)] -> [[Neuron]]

-- | Train the given neural network on the given samples using the
--   backpropagation algorithm using the given learning ratio (alpha) and
--   the given desired maximal bound for the global quadratic error on the
--   samples (epsilon)
trainU :: Double -> Double -> [[Neuron]] -> [(UArr Double, UArr Double)] -> [[Neuron]]

-- | Train the given neural network on the given samples using the
--   backpropagation algorithm using the given learning ratio (alpha) and
--   the given desired maximal bound for the global quadratic error on the
--   samples (epsilon)
train :: Double -> Double -> [[Neuron]] -> [([Double], [Double])] -> [[Neuron]]