Copyright	(c) Alexey Kuleshevich 2018-2022
License	BSD3
Maintainer	Alexey Kuleshevich <lehins@yandex.ru>
Stability	experimental
Portability	non-portable
Safe Haskell	Safe-Inferred
Language	Haskell2010

Data.Massiv.Array.Delayed

Contents

Delayed

Description

Synopsis

data D = D
delay :: (Index ix, Source r e) => Array r ix e -> Array D ix e
liftArray2' :: (HasCallStack, Index ix, Source r1 a, Source r2 b) => (a -> b -> e) -> Array r1 ix a -> Array r2 ix b -> Array D ix e
liftArray2M :: (Index ix, Source r1 a, Source r2 b, MonadThrow m) => (a -> b -> e) -> Array r1 ix a -> Array r2 ix b -> m (Array D ix e)
data DL = DL
toLoadArray :: forall r ix e. (Size r, Load r ix e) => Array r ix e -> Array DL ix e
makeLoadArrayS :: forall ix e. Index ix => Sz ix -> e -> (forall m. Monad m => (ix -> e -> m Bool) -> m ()) -> Array DL ix e
makeLoadArray :: forall ix e. Index ix => Comp -> Sz ix -> e -> (forall s. Scheduler s () -> (ix -> e -> ST s Bool) -> ST s ()) -> Array DL ix e
fromStrideLoad :: forall r ix e. StrideLoad r ix e => Stride ix -> Array r ix e -> Array DL ix e
data DS = DS
toStreamArray :: (Index ix, Source r e) => Array r ix e -> Vector DS e
toSteps :: Vector DS e -> Steps Id e
fromSteps :: Steps Id e -> Vector DS e
data DI = DI
toInterleaved :: (Index ix, Source r e) => Array r ix e -> Array DI ix e
fromInterleaved :: Array DI ix e -> Array D ix e
data DW = DW
data Window ix e = Window {
- windowStart :: !ix
- windowSize :: !(Sz ix)
- windowIndex :: ix -> e
- windowUnrollIx2 :: !(Maybe Int)
}
insertWindow :: Index ix => Array D ix e -> Window ix e -> Array DW ix e
getWindow :: Array DW ix e -> Maybe (Window ix e)
dropWindow :: Array DW ix e -> Array D ix e
makeWindowedArray :: (Index ix, Source r e) => Array r ix e -> ix -> Sz ix -> (ix -> e) -> Array DW ix e

Delayed

Delayed Pull Array

data D Source #

Delayed representation.

Constructors

D

Instances

Instances details

Show D Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods showsPrec :: Int -> D -> ShowS # show :: D -> String # showList :: [D] -> ShowS #
Size D Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods size :: Array D ix e -> Sz ix Source # unsafeResize :: (Index ix, Index ix') => Sz ix' -> Array D ix e -> Array D ix' e Source #
Strategy D Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods setComp :: Comp -> Array D ix e -> Array D ix e Source # getComp :: Array D ix e -> Comp Source # repr :: D
Index ix => Shape D ix Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods linearSizeHint :: Array D ix e -> LengthHint Source # linearSize :: Array D ix e -> Sz1 Source # outerSize :: Array D ix e -> Sz ix Source # maxLinearSize :: Array D ix e -> Maybe Sz1 Source # isNull :: Array D ix e -> Bool Source #
Source D e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods unsafeIndex :: Index ix => Array D ix e -> ix -> e Source # unsafeLinearIndex :: Index ix => Array D ix e -> Int -> e Source # unsafePrefIndex :: Index ix => Array D ix e -> PrefIndex ix e Source # unsafeOuterSlice :: (Index ix, Index (Lower ix)) => Array D ix e -> Sz (Lower ix) -> Int -> Array D (Lower ix) e Source # unsafeLinearSlice :: Index ix => Ix1 -> Sz1 -> Array D ix e -> Array D Ix1 e Source #
Num e => FoldNumeric D e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods sumArray :: Index ix => Array D ix e -> e Source # productArray :: Index ix => Array D ix e -> e Source # powerSumArray :: Index ix => Array D ix e -> Int -> e Source # unsafeDotProduct :: Index ix => Array D ix e -> Array D ix e -> e Source # foldArray :: Index ix => (e -> e -> e) -> e -> Array D ix e -> e Source #
Num e => Numeric D e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods plusScalar :: Index ix => Array D ix e -> e -> Array D ix e Source # minusScalar :: Index ix => Array D ix e -> e -> Array D ix e Source # scalarMinus :: Index ix => e -> Array D ix e -> Array D ix e Source # multiplyScalar :: Index ix => Array D ix e -> e -> Array D ix e Source # absPointwise :: Index ix => Array D ix e -> Array D ix e Source # additionPointwise :: Index ix => Array D ix e -> Array D ix e -> Array D ix e Source # subtractionPointwise :: Index ix => Array D ix e -> Array D ix e -> Array D ix e Source # multiplicationPointwise :: Index ix => Array D ix e -> Array D ix e -> Array D ix e Source # powerPointwise :: Index ix => Array D ix e -> Int -> Array D ix e Source # unsafeLiftArray :: Index ix => (e -> e) -> Array D ix e -> Array D ix e Source # unsafeLiftArray2 :: Index ix => (e -> e -> e) -> Array D ix e -> Array D ix e -> Array D ix e Source #
Floating e => NumericFloat D e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods divideScalar :: Index ix => Array D ix e -> e -> Array D ix e Source # scalarDivide :: Index ix => e -> Array D ix e -> Array D ix e Source # divisionPointwise :: Index ix => Array D ix e -> Array D ix e -> Array D ix e Source # recipPointwise :: Index ix => Array D ix e -> Array D ix e Source # sqrtPointwise :: Index ix => Array D ix e -> Array D ix e Source #
Index ix => Load D ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods makeArray :: Comp -> Sz ix -> (ix -> e) -> Array D ix e Source # makeArrayLinear :: Comp -> Sz ix -> (Int -> e) -> Array D ix e Source # replicate :: Comp -> Sz ix -> e -> Array D ix e Source # iterArrayLinearST_ :: Scheduler s () -> Array D ix e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array D ix e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array D ix e -> ST s (MArray s r' ix e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array D ix e -> IO (MArray RealWorld r' ix e) Source #
Index ix => Stream D ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods toStream :: Array D ix e -> Steps Id e Source # toStreamIx :: Array D ix e -> Steps Id (ix, e) Source #
Index ix => StrideLoad D ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods iterArrayLinearWithStrideST_ :: Scheduler s () -> Stride ix -> Sz ix -> Array D ix e -> (Int -> e -> ST s ()) -> ST s () Source #
Index ix => Foldable (Array D ix) Source #	Row-major sequential folding over a Delayed array.
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods fold :: Monoid m => Array D ix m -> m # foldMap :: Monoid m => (a -> m) -> Array D ix a -> m # foldMap' :: Monoid m => (a -> m) -> Array D ix a -> m # foldr :: (a -> b -> b) -> b -> Array D ix a -> b # foldr' :: (a -> b -> b) -> b -> Array D ix a -> b # foldl :: (b -> a -> b) -> b -> Array D ix a -> b # foldl' :: (b -> a -> b) -> b -> Array D ix a -> b # foldr1 :: (a -> a -> a) -> Array D ix a -> a # foldl1 :: (a -> a -> a) -> Array D ix a -> a # toList :: Array D ix a -> [a] # null :: Array D ix a -> Bool # length :: Array D ix a -> Int # elem :: Eq a => a -> Array D ix a -> Bool # maximum :: Ord a => Array D ix a -> a # minimum :: Ord a => Array D ix a -> a # sum :: Num a => Array D ix a -> a # product :: Num a => Array D ix a -> a #
Index ix => Applicative (Array D ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods pure :: a -> Array D ix a # (<>) :: Array D ix (a -> b) -> Array D ix a -> Array D ix b # liftA2 :: (a -> b -> c) -> Array D ix a -> Array D ix b -> Array D ix c # (>) :: Array D ix a -> Array D ix b -> Array D ix b # (<*) :: Array D ix a -> Array D ix b -> Array D ix a #
Functor (Array D ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods fmap :: (a -> b) -> Array D ix a -> Array D ix b # (<$) :: a -> Array D ix b -> Array D ix a #
(Ragged L ix e, Show e) => Show (Array D ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods showsPrec :: Int -> Array D ix e -> ShowS # show :: Array D ix e -> String # showList :: [Array D ix e] -> ShowS #
(Eq e, Index ix) => Eq (Array D ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods (==) :: Array D ix e -> Array D ix e -> Bool # (/=) :: Array D ix e -> Array D ix e -> Bool #
(Ord e, Index ix) => Ord (Array D ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull Methods compare :: Array D ix e -> Array D ix e -> Ordering # (<) :: Array D ix e -> Array D ix e -> Bool # (<=) :: Array D ix e -> Array D ix e -> Bool # (>) :: Array D ix e -> Array D ix e -> Bool # (>=) :: Array D ix e -> Array D ix e -> Bool # max :: Array D ix e -> Array D ix e -> Array D ix e # min :: Array D ix e -> Array D ix e -> Array D ix e #
data Array D ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Pull data Array D ix e = DArray { dComp :: !Comp dSize :: !(Sz ix) dPrefIndex :: !(PrefIndex ix e) }

delay :: (Index ix, Source r e) => Array r ix e -> Array D ix e Source #

O(1) Conversion from a source array to D representation.

liftArray2' :: (HasCallStack, Index ix, Source r1 a, Source r2 b) => (a -> b -> e) -> Array r1 ix a -> Array r2 ix b -> Array D ix e Source #

Same as liftArray2M, but throws an imprecise exception on mismatched sizes.

Since: 1.0.0

liftArray2M :: (Index ix, Source r1 a, Source r2 b, MonadThrow m) => (a -> b -> e) -> Array r1 ix a -> Array r2 ix b -> m (Array D ix e) Source #

Similar to zipWith, except dimensions of both arrays have to be the same, otherwise it throws SizeMismatchException.

Since: 1.0.0

Delayed Push Array

data DL Source #

Delayed load representation. Also known as Push array.

Constructors

DL

Instances

Instances details

Show DL Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods showsPrec :: Int -> DL -> ShowS # show :: DL -> String # showList :: [DL] -> ShowS #
Size DL Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods size :: Array DL ix e -> Sz ix Source # unsafeResize :: (Index ix, Index ix') => Sz ix' -> Array DL ix e -> Array DL ix' e Source #
Strategy DL Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods setComp :: Comp -> Array DL ix e -> Array DL ix e Source # getComp :: Array DL ix e -> Comp Source # repr :: DL
Index ix => Shape DL ix Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods linearSizeHint :: Array DL ix e -> LengthHint Source # linearSize :: Array DL ix e -> Sz1 Source # outerSize :: Array DL ix e -> Sz ix Source # maxLinearSize :: Array DL ix e -> Maybe Sz1 Source # isNull :: Array DL ix e -> Bool Source #
Index ix => Load DL ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods makeArray :: Comp -> Sz ix -> (ix -> e) -> Array DL ix e Source # makeArrayLinear :: Comp -> Sz ix -> (Int -> e) -> Array DL ix e Source # replicate :: Comp -> Sz ix -> e -> Array DL ix e Source # iterArrayLinearST_ :: Scheduler s () -> Array DL ix e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array DL ix e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array DL ix e -> ST s (MArray s r' ix e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array DL ix e -> IO (MArray RealWorld r' ix e) Source #
Index ix => Functor (Array DL ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods fmap :: (a -> b) -> Array DL ix a -> Array DL ix b # (<$) :: a -> Array DL ix b -> Array DL ix a #
Monoid (Array DL Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods mempty :: Array DL Ix1 e # mappend :: Array DL Ix1 e -> Array DL Ix1 e -> Array DL Ix1 e # mconcat :: [Array DL Ix1 e] -> Array DL Ix1 e #
Semigroup (Array DL Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push Methods (<>) :: Array DL Ix1 e -> Array DL Ix1 e -> Array DL Ix1 e # sconcat :: NonEmpty (Array DL Ix1 e) -> Array DL Ix1 e # stimes :: Integral b => b -> Array DL Ix1 e -> Array DL Ix1 e #
(Ragged L ix e, Show e) => Show (Array DL ix e) Source #
Instance details Defined in Data.Massiv.Array.Manifest.Boxed Methods showsPrec :: Int -> Array DL ix e -> ShowS # show :: Array DL ix e -> String # showList :: [Array DL ix e] -> ShowS #
data Array DL ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Push data Array DL ix e = DLArray { dlComp :: !Comp dlSize :: !(Sz ix) dlLoad :: Loader e }

toLoadArray :: forall r ix e. (Size r, Load r ix e) => Array r ix e -> Array DL ix e Source #

Convert any Loadable array into DL representation.

Since: 0.3.0

makeLoadArrayS Source #

Arguments

:: forall ix e. Index ix
=> Sz ix	Size of the resulting array
-> e	Default value to use for all cells that might have been ommitted by the writing function
-> (forall m. Monad m => (ix -> e -> m Bool) -> m ())	Writing function that described which elements to write into the target array.
-> Array DL ix e

Describe how an array should be loaded into memory sequentially. For parallelizable version see makeLoadArray.

Since: 0.3.1

makeLoadArray Source #

Arguments

:: forall ix e. Index ix
=> Comp	Computation strategy to use. Directly affects the scheduler that gets created for the loading function.
-> Sz ix	Size of the resulting array
-> e	Default value to use for all cells that might have been ommitted by the writing function
-> (forall s. Scheduler s () -> (ix -> e -> ST s Bool) -> ST s ())	Writing function that described which elements to write into the target array. It accepts a scheduler, that can be used for parallelization, as well as a safe element writing function.
-> Array DL ix e

Specify how an array should be loaded into memory. Unlike makeLoadArrayS, loading function accepts a scheduler, thus can be parallelized. If you need an unsafe version of this function see unsafeMakeLoadArray.

Since: 0.4.0

fromStrideLoad :: forall r ix e. StrideLoad r ix e => Stride ix -> Array r ix e -> Array DL ix e Source #

Convert an array that can be loaded with stride into DL representation.

Since: 0.3.0

Delayed Stream Array

data DS Source #

Delayed stream array that represents a sequence of values that can be loaded sequentially. Important distinction from other arrays is that its size might no be known until it is computed.

Constructors

DS

Instances

Instances details

Strategy DS Source #	For now only `Seq` strategy.
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods setComp :: Comp -> Array DS ix e -> Array DS ix e Source # getComp :: Array DS ix e -> Comp Source # repr :: DS
Shape DS Ix1 Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods linearSizeHint :: Array DS Ix1 e -> LengthHint Source # linearSize :: Array DS Ix1 e -> Sz1 Source # outerSize :: Array DS Ix1 e -> Sz Ix1 Source # maxLinearSize :: Array DS Ix1 e -> Maybe Sz1 Source # isNull :: Array DS Ix1 e -> Bool Source #
Load DS Ix1 e Source #	O(n) - `size` implementation.
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods makeArray :: Comp -> Sz Ix1 -> (Ix1 -> e) -> Array DS Ix1 e Source # makeArrayLinear :: Comp -> Sz Ix1 -> (Int -> e) -> Array DS Ix1 e Source # replicate :: Comp -> Sz Ix1 -> e -> Array DS Ix1 e Source # iterArrayLinearST_ :: Scheduler s () -> Array DS Ix1 e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array DS Ix1 e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array DS Ix1 e -> ST s (MArray s r' Ix1 e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array DS Ix1 e -> IO (MArray RealWorld r' Ix1 e) Source #
Stream DS Ix1 e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods toStream :: Array DS Ix1 e -> Steps Id e Source # toStreamIx :: Array DS Ix1 e -> Steps Id (Ix1, e) Source #
Foldable (Array DS Ix1) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods fold :: Monoid m => Array DS Ix1 m -> m # foldMap :: Monoid m => (a -> m) -> Array DS Ix1 a -> m # foldMap' :: Monoid m => (a -> m) -> Array DS Ix1 a -> m # foldr :: (a -> b -> b) -> b -> Array DS Ix1 a -> b # foldr' :: (a -> b -> b) -> b -> Array DS Ix1 a -> b # foldl :: (b -> a -> b) -> b -> Array DS Ix1 a -> b # foldl' :: (b -> a -> b) -> b -> Array DS Ix1 a -> b # foldr1 :: (a -> a -> a) -> Array DS Ix1 a -> a # foldl1 :: (a -> a -> a) -> Array DS Ix1 a -> a # toList :: Array DS Ix1 a -> [a] # null :: Array DS Ix1 a -> Bool # length :: Array DS Ix1 a -> Int # elem :: Eq a => a -> Array DS Ix1 a -> Bool # maximum :: Ord a => Array DS Ix1 a -> a # minimum :: Ord a => Array DS Ix1 a -> a # sum :: Num a => Array DS Ix1 a -> a # product :: Num a => Array DS Ix1 a -> a #
Applicative (Array DS Ix1) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods pure :: a -> Array DS Ix1 a # (<>) :: Array DS Ix1 (a -> b) -> Array DS Ix1 a -> Array DS Ix1 b # liftA2 :: (a -> b -> c) -> Array DS Ix1 a -> Array DS Ix1 b -> Array DS Ix1 c # (>) :: Array DS Ix1 a -> Array DS Ix1 b -> Array DS Ix1 b # (<*) :: Array DS Ix1 a -> Array DS Ix1 b -> Array DS Ix1 a #
Functor (Array DS Ix1) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods fmap :: (a -> b) -> Array DS Ix1 a -> Array DS Ix1 b # (<$) :: a -> Array DS Ix1 b -> Array DS Ix1 a #
Monad (Array DS Ix1) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods (>>=) :: Array DS Ix1 a -> (a -> Array DS Ix1 b) -> Array DS Ix1 b # (>>) :: Array DS Ix1 a -> Array DS Ix1 b -> Array DS Ix1 b # return :: a -> Array DS Ix1 a #
Monoid (Array DS Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods mempty :: Array DS Ix1 e # mappend :: Array DS Ix1 e -> Array DS Ix1 e -> Array DS Ix1 e # mconcat :: [Array DS Ix1 e] -> Array DS Ix1 e #
Semigroup (Array DS Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Methods (<>) :: Array DS Ix1 e -> Array DS Ix1 e -> Array DS Ix1 e # sconcat :: NonEmpty (Array DS Ix1 e) -> Array DS Ix1 e # stimes :: Integral b => b -> Array DS Ix1 e -> Array DS Ix1 e #
IsList (Array DS Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream Associated Types type Item (Array DS Ix1 e) # Methods fromList :: [Item (Array DS Ix1 e)] -> Array DS Ix1 e # fromListN :: Int -> [Item (Array DS Ix1 e)] -> Array DS Ix1 e # toList :: Array DS Ix1 e -> [Item (Array DS Ix1 e)] #
Show e => Show (Array DS Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Manifest.Boxed Methods showsPrec :: Int -> Array DS Ix1 e -> ShowS # show :: Array DS Ix1 e -> String # showList :: [Array DS Ix1 e] -> ShowS #
newtype Array DS Ix1 e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream newtype Array DS Ix1 e = DSArray { dsArray :: Steps Id e }
type Item (Array DS Ix1 e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Stream type Item (Array DS Ix1 e) = e

toStreamArray :: (Index ix, Source r e) => Array r ix e -> Vector DS e Source #

Flatten an array into a stream of values.

Since: 0.4.1

toSteps :: Vector DS e -> Steps Id e Source #

O(1) - Convert delayed stream array into Steps.

Since: 0.4.1

fromSteps :: Steps Id e -> Vector DS e Source #

O(1) - Convert Steps into delayed stream array

Since: 0.4.1

Delayed Interleaved Array

data DI Source #

Delayed array that will be loaded in an interleaved fashion during parallel computation.

Warning - Will be deprecated in the next major version update.

Constructors

DI

Instances

Instances details

Size DI Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods size :: Array DI ix e -> Sz ix Source # unsafeResize :: (Index ix, Index ix') => Sz ix' -> Array DI ix e -> Array DI ix' e Source #
Strategy DI Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods setComp :: Comp -> Array DI ix e -> Array DI ix e Source # getComp :: Array DI ix e -> Comp Source # repr :: DI
Index ix => Shape DI ix Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods linearSizeHint :: Array DI ix e -> LengthHint Source # linearSize :: Array DI ix e -> Sz1 Source # outerSize :: Array DI ix e -> Sz ix Source # maxLinearSize :: Array DI ix e -> Maybe Sz1 Source # isNull :: Array DI ix e -> Bool Source #
Index ix => Load DI ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods makeArray :: Comp -> Sz ix -> (ix -> e) -> Array DI ix e Source # makeArrayLinear :: Comp -> Sz ix -> (Int -> e) -> Array DI ix e Source # replicate :: Comp -> Sz ix -> e -> Array DI ix e Source # iterArrayLinearST_ :: Scheduler s () -> Array DI ix e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array DI ix e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array DI ix e -> ST s (MArray s r' ix e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array DI ix e -> IO (MArray RealWorld r' ix e) Source #
Index ix => StrideLoad DI ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods iterArrayLinearWithStrideST_ :: Scheduler s () -> Stride ix -> Sz ix -> Array DI ix e -> (Int -> e -> ST s ()) -> ST s () Source #
Index ix => Foldable (Array DI ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods fold :: Monoid m => Array DI ix m -> m # foldMap :: Monoid m => (a -> m) -> Array DI ix a -> m # foldMap' :: Monoid m => (a -> m) -> Array DI ix a -> m # foldr :: (a -> b -> b) -> b -> Array DI ix a -> b # foldr' :: (a -> b -> b) -> b -> Array DI ix a -> b # foldl :: (b -> a -> b) -> b -> Array DI ix a -> b # foldl' :: (b -> a -> b) -> b -> Array DI ix a -> b # foldr1 :: (a -> a -> a) -> Array DI ix a -> a # foldl1 :: (a -> a -> a) -> Array DI ix a -> a # toList :: Array DI ix a -> [a] # null :: Array DI ix a -> Bool # length :: Array DI ix a -> Int # elem :: Eq a => a -> Array DI ix a -> Bool # maximum :: Ord a => Array DI ix a -> a # minimum :: Ord a => Array DI ix a -> a # sum :: Num a => Array DI ix a -> a # product :: Num a => Array DI ix a -> a #
Index ix => Applicative (Array DI ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods pure :: a -> Array DI ix a # (<>) :: Array DI ix (a -> b) -> Array DI ix a -> Array DI ix b # liftA2 :: (a -> b -> c) -> Array DI ix a -> Array DI ix b -> Array DI ix c # (>) :: Array DI ix a -> Array DI ix b -> Array DI ix b # (<*) :: Array DI ix a -> Array DI ix b -> Array DI ix a #
Functor (Array DI ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods fmap :: (a -> b) -> Array DI ix a -> Array DI ix b # (<$) :: a -> Array DI ix b -> Array DI ix a #
(Ragged L ix e, Show e) => Show (Array DI ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods showsPrec :: Int -> Array DI ix e -> ShowS # show :: Array DI ix e -> String # showList :: [Array DI ix e] -> ShowS #
(Index ix, Eq e) => Eq (Array DI ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods (==) :: Array DI ix e -> Array DI ix e -> Bool # (/=) :: Array DI ix e -> Array DI ix e -> Bool #
(Index ix, Ord e) => Ord (Array DI ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved Methods compare :: Array DI ix e -> Array DI ix e -> Ordering # (<) :: Array DI ix e -> Array DI ix e -> Bool # (<=) :: Array DI ix e -> Array DI ix e -> Bool # (>) :: Array DI ix e -> Array DI ix e -> Bool # (>=) :: Array DI ix e -> Array DI ix e -> Bool # max :: Array DI ix e -> Array DI ix e -> Array DI ix e # min :: Array DI ix e -> Array DI ix e -> Array DI ix e #
newtype Array DI ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Interleaved newtype Array DI ix e = DIArray { diArray :: Array D ix e }

toInterleaved :: (Index ix, Source r e) => Array r ix e -> Array DI ix e Source #

Convert a source array into an array that, when computed, will have its elemets evaluated out of order (interleaved amongst cores), hence making unbalanced computation better parallelizable.

fromInterleaved :: Array DI ix e -> Array D ix e Source #

O(1) - Unwrap the interleved array.

Since: 0.2.1

Delayed Windowed Array

data DW Source #

Delayed Windowed Array representation.

Constructors

DW

Instances

Instances details

Strategy DW Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods setComp :: Comp -> Array DW ix e -> Array DW ix e Source # getComp :: Array DW ix e -> Comp Source # repr :: DW
Index ix => Shape DW ix Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods linearSizeHint :: Array DW ix e -> LengthHint Source # linearSize :: Array DW ix e -> Sz1 Source # outerSize :: Array DW ix e -> Sz ix Source # maxLinearSize :: Array DW ix e -> Maybe Sz1 Source # isNull :: Array DW ix e -> Bool Source #
Load DW Ix1 e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods makeArray :: Comp -> Sz Ix1 -> (Ix1 -> e) -> Array DW Ix1 e Source # makeArrayLinear :: Comp -> Sz Ix1 -> (Int -> e) -> Array DW Ix1 e Source # replicate :: Comp -> Sz Ix1 -> e -> Array DW Ix1 e Source # iterArrayLinearST_ :: Scheduler s () -> Array DW Ix1 e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array DW Ix1 e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array DW Ix1 e -> ST s (MArray s r' Ix1 e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array DW Ix1 e -> IO (MArray RealWorld r' Ix1 e) Source #
Load DW Ix2 e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods makeArray :: Comp -> Sz Ix2 -> (Ix2 -> e) -> Array DW Ix2 e Source # makeArrayLinear :: Comp -> Sz Ix2 -> (Int -> e) -> Array DW Ix2 e Source # replicate :: Comp -> Sz Ix2 -> e -> Array DW Ix2 e Source # iterArrayLinearST_ :: Scheduler s () -> Array DW Ix2 e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array DW Ix2 e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array DW Ix2 e -> ST s (MArray s r' Ix2 e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array DW Ix2 e -> IO (MArray RealWorld r' Ix2 e) Source #
StrideLoad DW Ix1 e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods iterArrayLinearWithStrideST_ :: Scheduler s () -> Stride Ix1 -> Sz Ix1 -> Array DW Ix1 e -> (Int -> e -> ST s ()) -> ST s () Source #
StrideLoad DW Ix2 e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods iterArrayLinearWithStrideST_ :: Scheduler s () -> Stride Ix2 -> Sz Ix2 -> Array DW Ix2 e -> (Int -> e -> ST s ()) -> ST s () Source #
(Index (IxN n), Load DW (Ix (n - 1)) e) => Load DW (IxN n) e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods makeArray :: Comp -> Sz (IxN n) -> (IxN n -> e) -> Array DW (IxN n) e Source # makeArrayLinear :: Comp -> Sz (IxN n) -> (Int -> e) -> Array DW (IxN n) e Source # replicate :: Comp -> Sz (IxN n) -> e -> Array DW (IxN n) e Source # iterArrayLinearST_ :: Scheduler s () -> Array DW (IxN n) e -> (Int -> e -> ST s ()) -> ST s () Source # iterArrayLinearWithSetST_ :: Scheduler s () -> Array DW (IxN n) e -> (Ix1 -> e -> ST s ()) -> (Ix1 -> Sz1 -> e -> ST s ()) -> ST s () Source # unsafeLoadIntoST :: Manifest r' e => MVector s r' e -> Array DW (IxN n) e -> ST s (MArray s r' (IxN n) e) Source # unsafeLoadIntoIO :: Manifest r' e => MVector RealWorld r' e -> Array DW (IxN n) e -> IO (MArray RealWorld r' (IxN n) e) Source #
(Index (IxN n), StrideLoad DW (Ix (n - 1)) e) => StrideLoad DW (IxN n) e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods iterArrayLinearWithStrideST_ :: Scheduler s () -> Stride (IxN n) -> Sz (IxN n) -> Array DW (IxN n) e -> (Int -> e -> ST s ()) -> ST s () Source #
Functor (Array DW ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods fmap :: (a -> b) -> Array DW ix a -> Array DW ix b # (<$) :: a -> Array DW ix b -> Array DW ix a #
(Ragged L ix e, Load DW ix e, Show e) => Show (Array DW ix e) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods showsPrec :: Int -> Array DW ix e -> ShowS # show :: Array DW ix e -> String # showList :: [Array DW ix e] -> ShowS #
data Array DW ix e Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed data Array DW ix e = DWArray { dwArray :: !(Array D ix e) dwWindow :: !(Maybe (Window ix e)) }

data Window ix e Source #

Constructors

Window
Fields windowStart :: !ix Index of where window will start at. windowSize :: !(Sz ix) Size of the window windowIndex :: ix -> e Indexing function for the window windowUnrollIx2 :: !(Maybe Int) Setting this value during stencil application improves cache utilization by unrolling the loop for Ix2 and higher dimensions. Has no affect on arrays with one dimension.

Instances

Instances details

Functor (Window ix) Source #
Instance details Defined in Data.Massiv.Array.Delayed.Windowed Methods fmap :: (a -> b) -> Window ix a -> Window ix b # (<$) :: a -> Window ix b -> Window ix a #

insertWindow Source #

Arguments

:: Index ix
=> Array D ix e	Source array that will have a window inserted into it
-> Window ix e	Window to place inside the delayed array
-> Array DW ix e

Inserts a Window into a delayed array while scaling the window down if it doesn't fit inside that array.

Since: 0.3.0

getWindow :: Array DW ix e -> Maybe (Window ix e) Source #

Get the Window from a windowed array.

Since: 0.2.1

dropWindow :: Array DW ix e -> Array D ix e Source #

Drop the Window from a windowed array.

Since: 0.3.0

makeWindowedArray Source #

Arguments

:: (Index ix, Source r e)
=> Array r ix e	Source array that will have a window inserted into it
-> ix	Start index for the window
-> Sz ix	Size of the window
-> (ix -> e)	Indexing function foto use inside window
-> Array DW ix e

Construct a delayed windowed array by supply a separate element producing function for the interior of an array. This is very usful for stencil mapping, where interior function does not perform boundary checks, thus significantly speeding up computation process.

Since: 0.1.3