GHC 8.4 compatibility

- `-XNoStarIsType` is new in GHC 8.6
- can't use `-XCPP` to conditionally enable `LANGUAGE` pragmas (implicit module declaration)
- use a conditional in the .cabal file, but this is active for all modules
- need to replace all `*` with `Type` and `import Data.Kind (Type)` (works in 8.4)

grenade.cabal

@@ -58,6 +58,8 @@ library
   if impl(ghc < 8.0)
     ghc-options:       -fno-warn-incomplete-patterns
 
+  if impl(ghc >= 8.6)
+    default-extensions: NoStarIsType
 
   exposed-modules:
                        Grenade

src/Grenade/Core/Layer.hs

@@ -41,6 +41,8 @@ import           Control.Monad.Random ( MonadRandom )
 
 import           Data.List ( foldl' )
 
+import           Data.Kind (Type)
+
 import           Grenade.Core.Shape
 import           Grenade.Core.LearningParameters
 
@@ -50,7 +52,7 @@ import           Grenade.Core.LearningParameters
 class UpdateLayer x where
   -- | The type for the gradient for this layer.
   --   Unit if there isn't a gradient to pass back.
-  type Gradient x :: *
+  type Gradient x :: Type
 
   -- | Update a layer with its gradient and learning parameters
   runUpdate       :: LearningParameters -> x -> Gradient x -> x
@@ -72,7 +74,7 @@ class UpdateLayer x => Layer x (i :: Shape) (o :: Shape) where
   -- | The Wengert tape for this layer. Includes all that is required
   --   to generate the back propagated gradients efficiently. As a
   --   default, `S i` is fine.
-  type Tape x i o :: *
+  type Tape x i o :: Type
 
   -- | Used in training and scoring. Take the input from the previous
   --   layer, and give the output from this layer.

src/Grenade/Core/Network.hs

@@ -36,6 +36,8 @@ import           Data.Singletons
 import           Data.Singletons.Prelude
 import           Data.Serialize
 
+import           Data.Kind (Type)
+
 import           Grenade.Core.Layer
 import           Grenade.Core.LearningParameters
 import           Grenade.Core.Shape
@@ -48,7 +50,7 @@ import           Grenade.Core.Shape
 --
 --   Can be considered to be a heterogeneous list of layers which are able to
 --   transform the data shapes of the network.
-data Network :: [*] -> [Shape] -> * where
+data Network :: [Type] -> [Shape] -> Type where
     NNil  :: SingI i
           => Network '[] '[i]
 
@@ -66,7 +68,7 @@ instance (Show x, Show (Network xs rs)) => Show (Network (x ': xs) (i ': rs)) wh
 -- | Gradient of a network.
 --
 --   Parameterised on the layers of the network.
-data Gradients :: [*] -> * where
+data Gradients :: [Type] -> Type where
    GNil  :: Gradients '[]
 
    (:/>) :: UpdateLayer x
@@ -77,7 +79,7 @@ data Gradients :: [*] -> * where
 -- | Wegnert Tape of a network.
 --
 --   Parameterised on the layers and shapes of the network.
-data Tapes :: [*] -> [Shape] -> * where
+data Tapes :: [Type] -> [Shape] -> Type where
    TNil  :: SingI i
          => Tapes '[] '[i]
 
@@ -152,7 +154,7 @@ applyUpdate _ NNil GNil
 
 -- | A network can easily be created by hand with (:~>), but an easy way to
 --   initialise a random network is with the randomNetwork.
-class CreatableNetwork (xs :: [*]) (ss :: [Shape]) where
+class CreatableNetwork (xs :: [Type]) (ss :: [Shape]) where
   -- | Create a network with randomly initialised weights.
   --
   --   Calls to this function will not compile if the type of the neural

src/Grenade/Core/Shape.hs

@@ -8,7 +8,6 @@
 {-# LANGUAGE FlexibleContexts      #-}
 {-# LANGUAGE ScopedTypeVariables   #-}
 {-# LANGUAGE RankNTypes            #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Core.Shape

src/Grenade/Layers/Concat.hs

@@ -8,7 +8,6 @@
 {-# LANGUAGE FlexibleInstances     #-}
 {-# LANGUAGE ScopedTypeVariables   #-}
 {-# LANGUAGE StandaloneDeriving    #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Layers.Concat

src/Grenade/Layers/Convolution.hs

@@ -8,7 +8,6 @@
 {-# LANGUAGE MultiParamTypeClasses #-}
 {-# LANGUAGE FlexibleInstances     #-}
 {-# LANGUAGE FlexibleContexts      #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Layers.Convolution

src/Grenade/Layers/Crop.hs

@@ -6,7 +6,6 @@
 {-# LANGUAGE TypeFamilies          #-}
 {-# LANGUAGE FlexibleContexts      #-}
 {-# LANGUAGE MultiParamTypeClasses #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Layers.Crop

src/Grenade/Layers/Deconvolution.hs

@@ -8,7 +8,6 @@
 {-# LANGUAGE MultiParamTypeClasses #-}
 {-# LANGUAGE FlexibleInstances     #-}
 {-# LANGUAGE FlexibleContexts      #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Layers.Deconvolution

src/Grenade/Layers/Merge.hs

@@ -23,13 +23,15 @@ import           Data.Serialize
 
 import           Data.Singletons
 
+import           Data.Kind (Type)
+
 import           Grenade.Core
 
 -- | A Merging layer.
 --
 -- Similar to Concat layer, except sums the activations instead of creating a larger
 -- shape.
-data Merge :: * -> * -> * where
+data Merge :: Type -> Type -> Type where
   Merge :: x -> y -> Merge x y
 
 instance (Show x, Show y) => Show (Merge x y) where

src/Grenade/Layers/Pad.hs

@@ -6,7 +6,6 @@
 {-# LANGUAGE TypeFamilies          #-}
 {-# LANGUAGE MultiParamTypeClasses #-}
 {-# LANGUAGE FlexibleContexts      #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Core.Pad

src/Grenade/Layers/Pooling.hs

@@ -7,7 +7,6 @@
 {-# LANGUAGE TypeFamilies          #-}
 {-# LANGUAGE MultiParamTypeClasses #-}
 {-# LANGUAGE FlexibleContexts      #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Core.Pooling

src/Grenade/Layers/Reshape.hs

@@ -3,7 +3,6 @@
 {-# LANGUAGE TypeFamilies          #-}
 {-# LANGUAGE MultiParamTypeClasses #-}
 {-# LANGUAGE FlexibleContexts      #-}
-{-# LANGUAGE NoStarIsType          #-}
 {-# LANGUAGE UndecidableInstances  #-}
 {-|
 Module      : Grenade.Layers.Reshape

src/Grenade/Recurrent/Core/Layer.hs

@@ -8,6 +8,8 @@ module Grenade.Recurrent.Core.Layer (
   , RecurrentUpdateLayer (..)
   ) where
 
+import           Data.Kind (Type)
+
 import           Grenade.Core
 
 -- | Class for a recurrent layer.
@@ -15,11 +17,11 @@ import           Grenade.Core
 --   of an extra recurrent data shape.
 class UpdateLayer x => RecurrentUpdateLayer x where
   -- | Shape of data that is passed between each subsequent run of the layer
-  type RecurrentShape x   :: *
+  type RecurrentShape x   :: Type
 
 class (RecurrentUpdateLayer x, Num (RecurrentShape x)) => RecurrentLayer x (i :: Shape) (o :: Shape) where
   -- | Wengert Tape
-  type RecTape x i o :: *
+  type RecTape x i o :: Type
   -- | Used in training and scoring. Take the input from the previous
   --   layer, and give the output from this layer.
   runRecurrentForwards    :: x -> RecurrentShape x -> S i -> (RecTape x i o, RecurrentShape x, S o)

src/Grenade/Recurrent/Core/Network.hs

@@ -32,18 +32,20 @@ import           Data.Singletons ( SingI )
 import           Data.Singletons.Prelude ( Head, Last )
 import           Data.Serialize
 
+import           Data.Kind (Type)
+
 import           Grenade.Core
 import           Grenade.Recurrent.Core.Layer
 
 -- | Witness type to say indicate we're building up with a normal feed
 --   forward layer.
-data FeedForward :: * -> *
+data FeedForward :: Type -> Type
 -- | Witness type to say indicate we're building up with a recurrent layer.
-data Recurrent :: * -> *
+data Recurrent :: Type -> Type
 
 -- | Type of a recurrent neural network.
 --
---   The [*] type specifies the types of the layers.
+--   The [Type] type specifies the types of the layers.
 --
 --   The [Shape] type specifies the shapes of data passed between the layers.
 --
@@ -52,7 +54,7 @@ data Recurrent :: * -> *
 --
 --   Often, to make the definitions more concise, one will use a type alias
 --   for these empty data types.
-data RecurrentNetwork :: [*] -> [Shape] -> * where
+data RecurrentNetwork :: [Type] -> [Shape] -> Type where
   RNil   :: SingI i
          => RecurrentNetwork '[] '[i]
 
@@ -71,7 +73,7 @@ infixr 5 :~@>
 -- | Gradient of a network.
 --
 --   Parameterised on the layers of the network.
-data RecurrentGradient :: [*] -> * where
+data RecurrentGradient :: [Type] -> Type where
    RGNil  :: RecurrentGradient '[]
 
    (://>) :: UpdateLayer x
@@ -81,7 +83,7 @@ data RecurrentGradient :: [*] -> * where
 
 -- | Recurrent inputs (sideways shapes on an imaginary unrolled graph)
 --   Parameterised on the layers of a Network.
-data RecurrentInputs :: [*] -> * where
+data RecurrentInputs :: [Type] -> Type where
    RINil   :: RecurrentInputs '[]
 
    (:~~+>) :: (UpdateLayer x, Fractional (RecurrentInputs xs))
@@ -95,7 +97,7 @@ data RecurrentInputs :: [*] -> * where
 --
 --   We index on the time step length as well, to ensure
 --   that that all Tape lengths are the same.
-data RecurrentTape :: [*] -> [Shape] -> * where
+data RecurrentTape :: [Type] -> [Shape] -> Type where
    TRNil  :: SingI i
           => RecurrentTape '[] '[i]
 
@@ -204,7 +206,7 @@ instance (Show x, Show (RecurrentNetwork xs rs)) => Show (RecurrentNetwork (Recu
 
 -- | A network can easily be created by hand with (:~~>) and (:~@>), but an easy way to initialise a random
 --   recurrent network and a set of random inputs for it is with the randomRecurrent.
-class CreatableRecurrent (xs :: [*]) (ss :: [Shape]) where
+class CreatableRecurrent (xs :: [Type]) (ss :: [Shape]) where
   -- | Create a network of the types requested
   randomRecurrent :: MonadRandom m => m (RecurrentNetwork xs ss)
 

src/Grenade/Recurrent/Layers/BasicRecurrent.hs

@@ -18,6 +18,8 @@ import           Control.Monad.Random ( MonadRandom, getRandom )
 
 import           Data.Singletons.TypeLits
 
+import           Data.Kind (Type)
+
 import           Numeric.LinearAlgebra.Static
 
 import           GHC.TypeLits
@@ -27,7 +29,7 @@ import           Grenade.Recurrent.Core
 
 data BasicRecurrent :: Nat -- Input layer size
                     -> Nat -- Output layer size
-                    -> * where
+                    -> Type where
   BasicRecurrent :: ( KnownNat input
                     , KnownNat output
                     , KnownNat matrixCols
@@ -40,7 +42,7 @@ data BasicRecurrent :: Nat -- Input layer size
 
 data BasicRecurrent' :: Nat -- Input layer size
                      -> Nat -- Output layer size
-                     -> * where
+                     -> Type where
   BasicRecurrent' :: ( KnownNat input
                      , KnownNat output
                      , KnownNat matrixCols

src/Grenade/Recurrent/Layers/ConcatRecurrent.hs

@@ -27,6 +27,8 @@ import           Data.Serialize
 import           Data.Singletons
 import           GHC.TypeLits
 
+import           Data.Kind (Type)
+
 import           Grenade.Core
 import           Grenade.Recurrent.Core
 
@@ -45,7 +47,7 @@ import           Numeric.LinearAlgebra.Static ( (#), split, R )
 --
 -- 3D images become 3D images with more channels. The sizes must be the same, one can use Pad
 -- and Crop layers to ensure this is the case.
-data ConcatRecurrent :: Shape -> * -> Shape -> * -> * where
+data ConcatRecurrent :: Shape -> Type -> Shape -> Type -> Type where
   ConcatRecLeft  :: x -> y -> ConcatRecurrent m (Recurrent x) n (FeedForward y)
   ConcatRecRight :: x -> y -> ConcatRecurrent m (FeedForward x) n (Recurrent y)
   ConcatRecBoth  :: x -> y -> ConcatRecurrent m (Recurrent x) n   (Recurrent y)

src/Grenade/Recurrent/Layers/LSTM.hs

@@ -23,6 +23,8 @@ import           Data.Proxy
 import           Data.Serialize
 import           Data.Singletons.TypeLits
 
+import           Data.Kind (Type)
+
 import qualified Numeric.LinearAlgebra as LA
 import           Numeric.LinearAlgebra.Static
 
@@ -36,14 +38,14 @@ import           Grenade.Layers.Internal.Update
 --   This is a Peephole formulation, so the recurrent shape is
 --   just the cell state, the previous output is not held or used
 --   at all.
-data LSTM :: Nat -> Nat -> * where
+data LSTM :: Nat -> Nat -> Type where
   LSTM :: ( KnownNat input
           , KnownNat output
           ) => !(LSTMWeights input output) -- Weights
             -> !(LSTMWeights input output) -- Momentums
             -> LSTM input output
 
-data LSTMWeights :: Nat -> Nat -> * where
+data LSTMWeights :: Nat -> Nat -> Type where
   LSTMWeights :: ( KnownNat input
                  , KnownNat output
                  ) => {