Data.Massiv.Vector

Documentation

type Vector r e = Array r Ix1 e

Type synonym for a single dimension array, or simply a flat vector.

Since: 0.5.0

type MVector s r e = MArray s r Ix1 e

Type synonym for a single dimension mutable array, or simply a flat mutable vector.

Since: 0.5.0

Accessors

Size

slength :: Stream r ix e => Array r ix e -> Maybe Sz1

O(1) - Get the length of a Stream array, but only if it is known exactly in constant time without looking at any of the elements in the array.

Related: maxSize, size, elemsCount and totalElem

Examples

>>> slength $ sfromList []
Nothing
>>> slength $ sreplicate 5 ()
Just (Sz1 5)
>>> slength $ makeArrayLinearR D Seq (Sz1 5) id
Just (Sz1 5)
>>> slength $ sunfoldr (\x -> Just (x, x)) (0 :: Int)
Nothing
>>> slength $ sunfoldrN 10 (\x -> Just (x, x)) (0 :: Int)
Nothing
>>> slength $ sunfoldrExactN 10 (\x -> (x, x)) (0 :: Int)
Just (Sz1 10)

Similar:

Data.Foldable.length

For some data structures, like a list for example, it is an O(n) operation, because there is a need to evaluate the full spine and possibly even the elements in order to get the full length. With Stream vectors that is rarely the case.

Data.Vector.Generic.length

In the vector package this function will always break fusion, unless it is the only operation that is applied to the vector.

Since: 0.5.0

maxSize :: Load r ix e => Array r ix e -> Maybe (Sz ix)

O(1) - Get the possible maximum size of an immutabe array. If the lookup of size in constant time is not possible, Nothing will be returned. This value will be used as the initial size of the mutable array into which the loading will happen.

Since: 0.5.0

size :: Load r ix e => Array r ix e -> Sz ix

Get the exact size of an immutabe array. Most of the time will produce the size in constant time, except for DS representation, which could result in evaluation of the whole stream. See maxSize and slength for more info.

Since: 0.1.0

snull :: Load r ix e => Array r ix e -> Bool

O(1) - Check whether a Stream array is empty or not. It only looks at the exact size (i.e. slength), if it is available, otherwise checks if there is at least one element in a stream.

Related: isEmpty, isNotEmpty

Examples

>>> snull sempty
True
>>> snull (empty :: Array D Ix5 Int)
True
>>> snull $ ssingleton "A Vector with a single String element"
False
>>> snull $ sfromList []
True
>>> snull $ sfromList [1 :: Int ..]
False

Similar:

Data.Foldable.null

List fusion is also broken with a check for emptiness, unless there are no other consumers of the list.

Data.Vector.Generic.null

Same as with length, unless it is the only operation applied to the vector it will break fusion and will result in the vector being fully materialized in memory.

Since: 0.5.0

Indexing

(!?) :: (Manifest r ix e, MonadThrow m) => Array r ix e -> ix -> m e infixl 4

O(1) - Infix version of indexM.

Exceptions: IndexOutOfBoundsException

Examples

>>> import Data.Massiv.Array as A
>>> :set -XTypeApplications
>>> a <- fromListsM @U @Ix2 @Int Seq [[1,2,3],[4,5,6]]
>>> a
Array U Seq (Sz (2 :. 3))
  [ [ 1, 2, 3 ]
  , [ 4, 5, 6 ]
  ]
>>> a !? 0 :. 2
3
>>> a !? 0 :. 3
*** Exception: IndexOutOfBoundsException: (0 :. 3) is not safe for (Sz (2 :. 3))
>>> a !? 0 :. 3 :: Maybe Int
Nothing

Since: 0.1.0

(!) :: Manifest r ix e => Array r ix e -> ix -> e infixl 4

O(1) - Infix version of index'.

Examples

>>> import Data.Massiv.Array as A
>>> a = computeAs U $ iterateN (Sz (2 :. 3)) succ (0 :: Int)
>>> a
Array U Seq (Sz (2 :. 3))
  [ [ 1, 2, 3 ]
  , [ 4, 5, 6 ]
  ]
>>> a ! 0 :. 2
3
>>> a ! 0 :. 3
*** Exception: IndexOutOfBoundsException: (0 :. 3) is not safe for (Sz (2 :. 3))

Since: 0.1.0

index :: Manifest r ix e => Array r ix e -> ix -> Maybe e

O(1) - Lookup an element in the array. Returns Nothing, when index is out of bounds and returns the element at the supplied index otherwise. Use indexM instead, since it is more general and it can just as well be used with Maybe.

Since: 0.1.0

index' :: Manifest r ix e => Array r ix e -> ix -> e

O(1) - Lookup an element in the array. This is a partial function and it can throw IndexOutOfBoundsException inside pure code. It is safer to use index instead.

Examples

>>> import Data.Massiv.Array
>>> :set -XOverloadedLists
>>> xs = [0..100] :: Array U Ix1 Int
>>> index' xs 50
50
>>> index' xs 150
*** Exception: IndexOutOfBoundsException: 150 is not safe for (Sz1 101)

Since: 0.1.0

head' :: Source r Ix1 e => Vector r e -> e

O(1) - Get the first element of a Source vector. Throws an error on empty.

Related: shead', headM, sheadM, unconsM.

Examples

>>> head' (Ix1 10 ..: 10000000000000)
10
>>> head' (Ix1 10 ..: 10)
*** Exception: SizeEmptyException: (Sz1 0) corresponds to an empty array

Similar:

Data.List.head

Also constant time and partial. Fusion is broken if there other consumers of the list.

Data.Vector.Generic.head

Also constant time and partial. Will cause materialization of the full vector if any other function is applied to the vector.

Since: 0.5.0

shead' :: Stream r Ix1 e => Vector r e -> e

O(1) - Get the first element of a Stream vector. Throws an error on empty.

Related: head', headM, sheadM, unconsM.

Examples

>>> shead' $ sunfoldr (\x -> Just (x, x)) (0 :: Int)
0
>>> x = shead' $ sunfoldr (\_ -> Nothing) (0 :: Int)
>>> print x
*** Exception: SizeEmptyException: (Sz1 0) corresponds to an empty array

Since: 0.5.0

last' :: Source r Ix1 e => Vector r e -> e

O(1) - Get the last element of a Source vector. Throws an error on empty.

Related: lastM, unsnocM

Examples

>>> last' (Ix1 10 ... 10000000000000)
10000000000000
>>> last' (fromList Seq [] :: Array P Ix1 Int)
*** Exception: SizeEmptyException: (Sz1 0) corresponds to an empty array

Similar:

Data.List.last

Also partial, but it has O(n) complixity. Fusion is broken if there other consumers of the list.

Data.Vector.Generic.last

Also constant time and partial. Will cause materialization of the full vector if any other function is applied to the vector.

Since: 0.5.0

Monadic Indexing

indexM :: (Manifest r ix e, MonadThrow m) => Array r ix e -> ix -> m e

O(1) - Lookup an element in the array.

Exceptions: IndexOutOfBoundsException

Since: 0.3.0

headM :: (Source r Ix1 e, MonadThrow m) => Vector r e -> m e

O(1) - Get the first element of a Source vector.

Related: head', shead', sheadM, unconsM.

Throws Exceptions: SizeEmptyException

Examples

>>> headM (Ix1 10 ..: 10000000000000)
10
>>> headM (Ix1 10 ..: 10000000000000) :: Maybe Int
Just 10
>>> headM (empty :: Array D Ix1 Int) :: Maybe Int
Nothing
>>> either show (const "") $ headM (Ix1 10 ..: 10)
"SizeEmptyException: (Sz1 0) corresponds to an empty array"

Similar:

Data.Maybe.listToMaybe

It also a safe way to get the head of the list, except it is restricted to Maybe

Since: 0.5.0

sheadM :: (Stream r Ix1 e, MonadThrow m) => Vector r e -> m e

O(1) - Get the first element of a Stream vector.

Related: head', shead', headM, unconsM.

Throws Exceptions: SizeEmptyException

Examples

>>> maybe 101 id $ sheadM (empty :: Vector D Int)
101
>>> maybe 101 id $ sheadM (singleton 202 :: Vector D Int)
202
>>> sheadM $ sunfoldr (\x -> Just (x, x)) (0 :: Int)
0
>>> x <- sheadM $ sunfoldr (\_ -> Nothing) (0 :: Int)
*** Exception: SizeEmptyException: (Sz1 0) corresponds to an empty array

Since: 0.5.0

lastM :: (Source r Ix1 e, MonadThrow m) => Vector r e -> m e

O(1) - Get the last element of a Source vector.

Related: last', unsnocM

Throws Exceptions: SizeEmptyException

Examples

>>> lastM (Ix1 10 ... 10000000000000)
10000000000000
>>> lastM (Ix1 10 ... 10000000000000) :: Maybe Int
Just 10000000000000
>>> either show (const "") $ lastM (fromList Seq [] :: Array P Ix1 Int)
"SizeEmptyException: (Sz1 0) corresponds to an empty array"

Since: 0.5.0

unconsM :: (MonadThrow m, Source r Ix1 e) => Vector r e -> m (e, Vector r e)

O(1) - Take one element off of the Source vector from the left side, as well as the remaining part of the vector in delayed D representation.

Related: head', shead', headM, sheadM, cons

Throws Exceptions: SizeEmptyException

Examples

>>> unconsM (fromList Seq [1,2,3] :: Array P Ix1 Int)
(1,Array P Seq (Sz1 2)
  [ 2, 3 ])

Similar:

Data.List.uncons

Same concept, except it is restricted to Maybe instead of the more general MonadThrow

Since: 0.3.0

unsnocM :: (MonadThrow m, Source r Ix1 e) => Vector r e -> m (Vector r e, e)

O(1) - Take one element off of the vector from the right side, as well as the remaining part of the vector.

Related: last', lastM, snoc

Throws Exceptions: SizeEmptyException

Examples

>>> unsnocM (fromList Seq [1,2,3] :: Array P Ix1 Int)
(Array P Seq (Sz1 2)
  [ 1, 2 ],3)

Since: 0.3.0

Slicing

slice :: Source r Ix1 e => Ix1 -> Sz1 -> Vector r e -> Vector r e

O(1) - Take a slice of a Source vector. Never fails, instead adjusts the indices.

Examples

>>> slice 10 5 (Ix1 0 ... 10000000000000)
Array D Seq (Sz1 5)
  [ 10, 11, 12, 13, 14 ]
>>> slice (-10) 5 (Ix1 0 ... 10000000000000)
Array D Seq (Sz1 5)
  [ 0, 1, 2, 3, 4 ]
>>> slice 9999999999998 50 (Ix1 0 ... 10000000000000)
Array D Seq (Sz1 3)
  [ 9999999999998, 9999999999999, 10000000000000 ]

Since: 0.5.0

slice' :: Source r Ix1 e => Ix1 -> Sz1 -> Vector r e -> Vector r e

O(1) - Take a slice of a Source vector. Throws an error on incorrect indices.

Examples

>>> slice' 10 5 (Ix1 0 ... 100)
Array D Seq (Sz1 5)
  [ 10, 11, 12, 13, 14 ]
>>> slice' (-10) 5 (Ix1 0 ... 100)
Array D *** Exception: SizeSubregionException: (Sz1 101) is to small for -10 (Sz1 5)
>>> slice' 98 50 (Ix1 0 ... 100)
Array D *** Exception: SizeSubregionException: (Sz1 101) is to small for 98 (Sz1 50)
>>> slice' 9999999999998 50 (Ix1 0 ... 10000000000000)
Array D *** Exception: SizeSubregionException: (Sz1 10000000000001) is to small for 9999999999998 (Sz1 50)
>>> slice' 9999999999998 3 (Ix1 0 ... 10000000000000)
Array D Seq (Sz1 3)
  [ 9999999999998, 9999999999999, 10000000000000 ]

Since: 0.5.0

sliceM :: (Source r Ix1 e, MonadThrow m) => Ix1 -> Sz1 -> Vector r e -> m (Vector r e)

O(1) - Take a slice of a Source vector. Throws an error on incorrect indices.

Throws Exceptions: SizeSubregionException

Examples

Since: 0.5.0

sslice :: Stream r Ix1 e => Ix1 -> Sz1 -> Vector r e -> Vector DS e

Take a slice of a Stream vector. Never fails, instead adjusts the indices.

Examples

>>> sslice 10 5 (Ix1 0 ... 10000000000000)
Array DS Seq (Sz1 5)
  [ 10, 11, 12, 13, 14 ]
>>> sslice 10 5 (sfromList [0 :: Int .. ])
Array DS Seq (Sz1 5)
  [ 10, 11, 12, 13, 14 ]
>>> sslice (-10) 5 (Ix1 0 ... 10000000000000)
Array DS Seq (Sz1 5)
  [ 0, 1, 2, 3, 4 ]

Unlike slice it has to iterate through each element until the staring index is reached, therefore something like sslice 9999999999998 50 (Ix1 0 ... 10000000000000) will not be feasable.

>>> import System.Timeout (timeout)
>>> let smallArr = sslice 9999999999998 50 (Ix1 0 ... 10000000000000)
>>> timeout 500000 (computeIO smallArr :: IO (Array P Ix1 Int))
Nothing

Since: 0.5.0

sliceAt :: Source r Ix1 e => Sz1 -> Vector r e -> (Vector r e, Vector r e)

Samel as sliceAt, except it never fails.

Examples

Since: 0.5.0

sliceAt' :: Source r Ix1 e => Sz1 -> Vector r e -> (Vector r e, Vector r e)

Same as splitAt', except for a flat vector.

Examples

Since: 0.5.0

sliceAtM :: (Source r Ix1 e, MonadThrow m) => Sz1 -> Vector r e -> m (Vector r e, Vector r e)

Same as splitAtM, except for a flat vector.

Examples

Since: 0.5.0

Init

init :: Source r Ix1 e => Vector r e -> Vector r e

O(1) - Get a vector without the last element. Never fails.

Examples

>>> init (0 ..: 10)
Array D Seq (Sz1 9)
  [ 0, 1, 2, 3, 4, 5, 6, 7, 8 ]
>>> init (empty :: Array D Ix1 Int)
Array D Seq (Sz1 0)
  [  ]

Since: 0.5.0

init' :: Source r Ix1 e => Vector r e -> Vector r e

O(1) - Get a vector without the last element. Throws an error on empty

Examples

>>> init' (0 ..: 10)
Array D Seq (Sz1 9)
  [ 0, 1, 2, 3, 4, 5, 6, 7, 8 ]
>>> init' (empty :: Array D Ix1 Int)
Array D *** Exception: SizeEmptyException: (Sz1 0) corresponds to an empty array

Since: 0.5.0

initM :: (Source r Ix1 e, MonadThrow m) => Vector r e -> m (Vector r e)

O(1) - Get a vector without the last element. Throws an error on empty

Examples

>>> initM (0 ..: 10)
Array D Seq (Sz1 9)
  [ 0, 1, 2, 3, 4, 5, 6, 7, 8 ]
>>> maybe 0 sum $ initM (0 ..: 10)
36
>>> maybe 0 sum $ initM (empty :: Array D Ix1 Int)
0

Since: 0.5.0

Tail

tail :: Source r Ix1 e => Vector r e -> Vector r e

O(1) - Get a vector without the first element. Never fails

Examples

>>> tail (0 ..: 10)
Array D Seq (Sz1 9)
  [ 1, 2, 3, 4, 5, 6, 7, 8, 9 ]
>>> tail (empty :: Array D Ix1 Int)
Array D Seq (Sz1 0)
  [  ]

Since: 0.5.0

tail' :: Source r Ix1 e => Vector r e -> Vector r e

O(1) - Get a vector without the first element. Throws an error on empty

Examples

λ> tail' (0 ..: 10) Array D Seq (Sz1 9) [ 1, 2, 3, 4, 5, 6, 7, 8, 9 ] λ> tail' (empty :: Array D Ix1 Int) Array D *** Exception: SizeEmptyException: (Sz1 0) corresponds to an empty array

Since: 0.5.0

tailM :: (Source r Ix1 e, MonadThrow m) => Vector r e -> m (Vector r e)

O(1) - Get the vector without the first element. Throws an error on empty

Examples

>>> tailM (0 ..: 10)
Array D Seq (Sz1 9)
  [ 1, 2, 3, 4, 5, 6, 7, 8, 9 ]
>>> maybe 0 sum $ tailM (0 ..: 10)
45
>>> maybe 0 sum $ tailM (empty :: Array D Ix1 Int)
0

Since: 0.5.0

Take

take :: Source r Ix1 e => Sz1 -> Vector r e -> Vector r e

O(1) - Get the vector with the first n elements. Never fails

Examples

>>> take 5 (0 ..: 10)
Array D Seq (Sz1 5)
  [ 0, 1, 2, 3, 4 ]
>>> take (-5) (0 ..: 10)
Array D Seq (Sz1 0)
  [  ]
>>> take 100 (0 ..: 10)
Array D Seq (Sz1 10)
  [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ]
>>> 

Since: 0.5.0

take' :: Source r Ix1 e => Sz1 -> Vector r e -> Vector r e

O(1) - Get the vector with the first n elements. Throws an error size is less than n.

Examples

>>> take' 0 (0 ..: 0)
Array D Seq (Sz1 0)
  [  ]
>>> take' 5 (0 ..: 10)
Array D Seq (Sz1 5)
  [ 0, 1, 2, 3, 4 ]
>>> take' 15 (0 ..: 10)
Array D *** Exception: SizeSubregionException: (Sz1 10) is to small for 0 (Sz1 15)

Since: 0.5.0

takeM :: (Source r Ix1 e, MonadThrow m) => Sz1 -> Vector r e -> m (Vector r e)

O(1) - Get the vector with the first n elements. Throws an error size is less than n

Examples

>>> takeM 5 (0 ..: 10)
Array D Seq (Sz1 5)
  [ 0, 1, 2, 3, 4 ]
>>> maybe 0 sum $ takeM 5 (0 ..: 10)
10
>>> maybe (-1) sum $ takeM 15 (0 ..: 10)
-1

Since: 0.5.0

stake :: Stream r Ix1 e => Sz1 -> Vector r e -> Vector DS e

O(1) - Get a Stream vector with the first n elements. Never fails

Examples

Since: 0.5.0

Drop

drop :: Source r Ix1 e => Sz1 -> Vector r e -> Vector r e

Examples

Since: 0.5.0

drop' :: Source r Ix1 e => Sz1 -> Vector r e -> Vector r e

Examples

Since: 0.5.0

dropM :: (Source r Ix1 e, MonadThrow m) => Sz1 -> Vector r e -> m (Vector r e)

Examples

Since: 0.5.0

sdrop :: Stream r Ix1 e => Sz1 -> Vector r e -> Vector DS e

Keep all but the first n elements from the delayed stream vector.

Examples

Since: 0.5.0

Construction

Initialization

empty :: forall r ix e. Construct r ix e => Array r ix e

Create an Array with no elements. By itself it is not particularly useful, but it serves as a nice base for constructing larger arrays.

Examples

>>> import Data.Massiv.Array as A
>>> :set -XTypeApplications
>>> xs = empty @DL @Ix1 @Double
>>> snoc (cons 4 (cons 5 xs)) 22
Array DL Seq (Sz1 3)
  [ 4.0, 5.0, 22.0 ]

Since: 0.3.0

sempty :: Vector DS e

Create an empty delayed stream vector

Examples

Since: 0.5.0

singleton

Arguments

:: Construct r ix e

=> e

The only element

-> Array r ix e

Create an Array with a single element.

Examples

>>> import Data.Massiv.Array as A
>>> singleton 7 :: Array D Ix4 Double
Array D Seq (Sz (1 :> 1 :> 1 :. 1))
  [ [ [ [ 7.0 ]
      ]
    ]
  ]

Instead of specifying type signature we could use TypeApplications

>>> :set -XTypeApplications
>>> singleton @U @Ix4 @Double 7
Array U Seq (Sz (1 :> 1 :> 1 :. 1))
  [ [ [ [ 7.0 ]
      ]
    ]
  ]

Since: 0.1.0

ssingleton :: e -> Vector DS e

Create a delayed stream vector with a single element

Examples

Since: 0.5.0

cons :: Load r Ix1 e => e -> Vector r e -> Vector DL e

O(1) - Add an element to the vector from the left side

Since: 0.3.0

snoc :: Load r Ix1 e => Vector r e -> e -> Vector DL e

O(1) - Add an element to the vector from the right side

Since: 0.3.0

replicate :: forall ix e. Index ix => Comp -> Sz ix -> e -> Array DL ix e

Replicate the same element

Since: 0.3.0

sreplicate :: Sz1 -> e -> Vector DS e

Replicate the same element n times

Examples

Since: 0.5.0

generate :: Comp -> Sz1 -> (Ix1 -> e) -> Vector D e

Create a delayed vector of length n with a function that maps an index to an element. Same as makeLinearArray

Examples

Since: 0.5.0

sgenerate :: Sz1 -> (Ix1 -> e) -> Vector DS e

Create a delayed stream vector of length n with a function that maps an index to an element. Same as makeLinearArray

Examples

Since: 0.5.0

siterate :: (e -> e) -> e -> Vector DS e

Create a delayed stream vector of infinite length by repeatedly applying a function to the initial value.

Examples

>>> stake 10 $ siterate succ 'a'
Array DS Seq (Sz1 10)
  [ 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j' ]

Since: 0.5.2

siterateN :: Sz1 -> (e -> e) -> e -> Vector DS e

Create a delayed stream vector of length n by repeatedly applying a function to the initial value.

Examples

>>> siterateN 10 succ 'a'
Array DS Seq (Sz1 10)
  [ 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j' ]

Since: 0.5.0

Monadic initialization

sreplicateM :: Monad m => Sz1 -> m e -> m (Vector DS e)

Create a vector by using the same monadic action n times

Examples

Since: 0.5.0

sgenerateM :: Monad m => Sz1 -> (Ix1 -> m e) -> m (Vector DS e)

Create a delayed stream vector of length n with a monadic action that from an index generates an element.

Examples

Since: 0.5.0

siterateNM :: Monad m => Sz1 -> (e -> m e) -> e -> m (Vector DS e)

Create a delayed stream vector of length n by repeatedly apply a monadic action to the initial value.

Examples

Since: 0.5.0

Unfolding

sunfoldr :: (s -> Maybe (e, s)) -> s -> Vector DS e

Right unfolding function. Useful when it is unknown ahead of time how many elements a vector will have.

Example

>>> import Data.Massiv.Array as A
>>> sunfoldr (\i -> if i < 9 then Just (i * i, i + 1) else Nothing) (0 :: Int)
Array DS Seq (Sz1 9)
  [ 0, 1, 4, 9, 16, 25, 36, 49, 64 ]

Since: 0.5.0

sunfoldrM :: Monad m => (s -> m (Maybe (e, s))) -> s -> m (Vector DS e)

O(n) - Same as unfoldr, but with monadic generating function.

Examples

>>> import Control.Monad (when, guard)
>>> sunfoldrM (\i -> when (i == 0) (Left "Zero denominator") >> Right (guard (i < 5) >> Just (100 `div` i, i + 1))) (-10 :: Int)
Left "Zero denominator"
>>> sunfoldrM (\i -> when (i == 0) (Left "Zero denominator") >> Right (guard (i < -5) >> Just (100 `div` i, i + 1))) (-10 :: Int)
Right (Array DS Seq (Sz1 5)
  [ -10, -12, -13, -15, -17 ]
)

Since: 0.5.0

sunfoldrN

Arguments

:: Sz1

n - maximum number of elements that the vector will have

-> (s -> Maybe (e, s))

Unfolding function. Stops when Nothing is returned or maximum number of elements is reached.

-> s

Inititial element.

-> Vector DS e

O(n) - Right unfolding function with at most n number of elements.

Example

>>> import Data.Massiv.Array as A
>>> sunfoldrN 9 (\i -> Just (i*i, i + 1)) (0 :: Int)
Array DS Seq (Sz1 9)
  [ 0, 1, 4, 9, 16, 25, 36, 49, 64 ]

Since: 0.5.0

sunfoldrNM :: Monad m => Sz1 -> (s -> m (Maybe (e, s))) -> s -> m (Vector DS e)

O(n) - Same as unfoldrN, but with monadic generating function.

Examples

>>> import Control.Monad (guard)
>>> sunfoldrNM 6 (\i -> print i >> pure (guard (i < 5) >> Just (i * i, i + 1))) (10 :: Int)
10
Array DS Seq (Sz1 0)
  [  ]
>>> sunfoldrNM 6 (\i -> print i >> pure (guard (i < 15) >> Just (i * i, i + 1))) (10 :: Int)
10
11
12
13
14
15
Array DS Seq (Sz1 5)
  [ 100, 121, 144, 169, 196 ]

Since: 0.5.0

sunfoldrExactN :: Sz1 -> (s -> (e, s)) -> s -> Vector DS e

O(n) - Similar to unfoldrN, except the length of the resulting vector will be exactly n

Examples

>>> sunfoldrExactN 10 (\i -> (i * i, i + 1)) (10 :: Int)
Array DS Seq (Sz1 10)
  [ 100, 121, 144, 169, 196, 225, 256, 289, 324, 361 ]

Since: 0.5.0

sunfoldrExactNM :: Monad m => Sz1 -> (s -> m (e, s)) -> s -> m (Vector DS e)

O(n) - Similar to unfoldrNM, except the length of the resulting vector will be exactly n

Examples

λ> sunfoldrExactNM 11 (i -> pure (100 div i, i + 1)) (-10 :: Int) Array DS *** Exception: divide by zero λ> sunfoldrExactNM 11 (i -> guard (i /= 0) >> Just (100 div i, i + 1)) (-10 :: Int) Nothing λ> sunfoldrExactNM 9 (i -> guard (i /= 0) >> Just (100 div i, i + 1)) (-10 :: Int) Just (Array DS Seq (Sz1 9) [ -10, -12, -13, -15, -17, -20, -25, -34, -50 ] )

Since: 0.5.0

Enumeration

(...) :: Index ix => ix -> ix -> Array D ix ix infix 4

Handy synonym for rangeInclusive Seq. Similar to .. for list.

>>> Ix1 4 ... 10
Array D Seq (Sz1 7)
  [ 4, 5, 6, 7, 8, 9, 10 ]

Since: 0.3.0

(..:) :: Index ix => ix -> ix -> Array D ix ix infix 4

Handy synonym for range Seq

>>> Ix1 4 ..: 10
Array D Seq (Sz1 6)
  [ 4, 5, 6, 7, 8, 9 ]

Since: 0.3.0

enumFromN

Arguments

:: Num e

=> Comp

-> e

x - start value

-> Sz1

n - length of resulting vector.

-> Vector D e

Same as enumFromStepN with step dx = 1.

Related: senumFromN, senumFromStepN, enumFromStepN, rangeSize, rangeStepSize, range

Examples

>>> import Data.Massiv.Array
>>> enumFromN Seq (5 :: Double) 3
Array D Seq (Sz1 3)
  [ 5.0, 6.0, 7.0 ]

Similar:

Prelude.enumFromTo

Very similar to [i .. i + n - 1], except that senumFromN is faster, but it only works for Num and not for Enum elements

Data.Vector.Generic.enumFromN

Since: 0.1.0

senumFromN

Arguments

:: Num e

=> e

x - starting number

-> Sz1

n - length of resulting vector

-> Vector DS e

O(n) - Enumerate from a starting number x exactly n times with a step 1.

Related: senumFromStepN, enumFromN, enumFromStepN, rangeSize, rangeStepSize, range, rangeStep

Examples

>>> senumFromN (10 :: Int) 9
Array DS Seq (Sz1 9)
  [ 10, 11, 12, 13, 14, 15, 16, 17, 18 ]

Similar:

Prelude.enumFromTo

Very similar to [x .. x + n - 1], except that senumFromN is faster and it only works for Num and not for Enum elements

Data.Vector.Generic.enumFromN

Uses exactly the same implementation underneath.

Since: 0.5.0

enumFromStepN

Arguments

:: Num e

=> Comp

-> e

x - start number

-> e

dx - step number

-> Sz1

n - length of resulting vector

-> Vector D e

Enumerate from a starting number x exactly n times with a custom step value dx. Unlike senumFromStepN, there is no dependency on neigboring elements therefore enumFromStepN is parallelizable.

Related: senumFromN, senumFromStepN, enumFromN, rangeSize, rangeStepSize, range, rangeStepM

Examples

>>> import Data.Massiv.Array
>>> enumFromStepN Seq 1 (0.1 :: Double) 5
Array D Seq (Sz1 5)
  [ 1.0, 1.1, 1.2, 1.3, 1.4 ]
>>> enumFromStepN Seq (-pi :: Float) (pi/4) 9
Array D Seq (Sz1 9)
  [ -3.1415927, -2.3561945, -1.5707964, -0.78539824, 0.0, 0.78539824, 1.5707963, 2.3561947, 3.1415927 ]

Similar:

Prelude.enumFrom

Similar to take n [x, x + dx ..], except that enumFromStepN is parallelizable and it only works for Num and not for Enum elements. Floating point value will be slightly different as well.

Data.Vector.Generic.enumFromStepN

Similar in the outcome, but very different in the way it works.

Since: 0.1.0

senumFromStepN

Arguments

:: Num e

=> e

x - starting number

-> e

dx - Step

-> Sz1

n - length of resulting vector

-> Vector DS e

O(n) - Enumerate from a starting number x exactly n times with a custom step value dx

Examples

>>> senumFromStepN (5 :: Int) 2 10
Array DS Seq (Sz1 10)
  [ 5, 7, 9, 11, 13, 15, 17, 19, 21, 23 ]

Similar:

Prelude.enumFrom

Just like take n [x, x + dx ..], except that senumFromN is faster and it only works for Num and not for Enum elements

Data.Vector.Generic.enumFromStepN

Uses exactly the same implementation underneath.

Since: 0.5.0

Concatenation

sappend :: (Stream r1 Ix1 e, Stream r2 Ix1 e) => Vector r1 e -> Vector r2 e -> Vector DS e

Append two vectors together

Related: appendM, appendOuterM,

Examples

λ> sappend (1 ..: 6) (senumFromStepN 6 (-1) 6) Array DS Seq (Sz1 11) [ 1, 2, 3, 4, 5, 6, 5, 4, 3, 2, 1 ]

Similar:

Data.Semigroup.<>

DS and DL arrays have instances for Semigroup, so they will work in a similar fashion. sappend differs in that it accepts Stream arrays with possibly different representations.

Data.List.++

Same operation, but for lists.

Data.Vector.Generic.++

Uses exactly the same implementation underneath as sappend, except that it cannot append two vectors with different memory representations.

Since: 0.5.0

sconcat :: Stream r Ix1 e => [Vector r e] -> Vector DS e

Concat vectors together

Related: concatM, concatOuterM,

Examples

>>> sconcat [2 ... 6, empty, singleton 1, generate Seq 5 id]
Array DS Seq (Sz1 11)
  [ 2, 3, 4, 5, 6, 1, 0, 1, 2, 3, 4 ]
>>> sconcat [senumFromN 2 5, sempty, ssingleton 1, sgenerate 5 id]
Array DS Seq (Sz1 11)
  [ 2, 3, 4, 5, 6, 1, 0, 1, 2, 3, 4 ]

Similar:

Data.Monoid.mconcat

DS and DL arrays have instances for Monoid, so they will work in a similar fashion. sconcat differs in that it accepts Stream arrays of other representations.

Data.List.concat

Same operation, but for lists.

Data.Vector.Generic.concat

Uses exactly the same implementation underneath as sconcat.

Since: 0.5.0

smap :: Stream r ix a => (a -> b) -> Array r ix a -> Vector DS b

Map a function over a stream vector

Examples

Since: 0.5.0

simap :: Stream r ix a => (ix -> a -> b) -> Array r ix a -> Vector DS b

Map an index aware function over a stream vector

Examples

Since: 0.5.0

Monadic mapping

straverse :: (Stream r ix a, Applicative f) => (a -> f b) -> Array r ix a -> f (Vector DS b)

Traverse a stream vector with an applicative function.

Examples

Since: 0.5.0

sitraverse :: (Stream r ix a, Applicative f) => (ix -> a -> f b) -> Array r ix a -> f (Vector DS b)

Traverse a stream vector with an index aware applicative function.

Examples

Since: 0.5.0

smapM :: (Stream r ix a, Monad m) => (a -> m b) -> Array r ix a -> m (Vector DS b)

Traverse a stream vector with a monadic function.

Examples

Since: 0.5.0

smapM_ :: (Stream r ix a, Monad m) => (a -> m b) -> Array r ix a -> m ()

Traverse a stream vector with a monadic function, while discarding the result

Examples

Since: 0.5.0

simapM :: (Stream r ix a, Monad m) => (ix -> a -> m b) -> Array r ix a -> m (Vector DS b)

Traverse a stream vector with a monadic index aware function.

Corresponds to: mapM (uncurry f) . imap (,) v

Examples

Since: 0.5.0

simapM_ :: (Stream r ix a, Monad m) => (ix -> a -> m b) -> Array r ix a -> m ()

Traverse a stream vector with a monadic index aware function, while discarding the result

Examples

Since: 0.5.0

sforM :: (Stream r ix a, Monad m) => Array r ix a -> (a -> m b) -> m (Vector DS b)

Same as smapM, but with arguments flipped.

Examples

Since: 0.5.0

sforM_ :: (Stream r ix a, Monad m) => Array r ix a -> (a -> m b) -> m ()

Same as smapM_, but with arguments flipped.

Examples

Since: 0.5.0

siforM :: (Stream r ix a, Monad m) => Array r ix a -> (ix -> a -> m b) -> m (Vector DS b)

Same as simapM, but with arguments flipped.

Examples

Since: 0.5.0

siforM_ :: (Stream r ix a, Monad m) => Array r ix a -> (ix -> a -> m b) -> m ()

Same as simapM_, but with arguments flipped.

Examples

Since: 0.5.0

Zipping

szip :: (Stream ra Ix1 a, Stream rb Ix1 b) => Vector ra a -> Vector rb b -> Vector DS (a, b)

Zip two arrays in a row-major order together together into a flat vector. Resulting length of a vector will be the smallest number of elements of the supplied arrays.

Examples

Since: 0.5.0

szip3 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c) => Vector ra a -> Vector rb b -> Vector rc c -> Vector DS (a, b, c)

Since: 0.5.0

szip4 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d) => Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector DS (a, b, c, d)

Since: 0.5.0

szip5 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e) => Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector DS (a, b, c, d, e)

Since: 0.5.0

szip6 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f) => Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> Vector DS (a, b, c, d, e, f)

Since: 0.5.0

szipWith :: (Stream ra Ix1 a, Stream rb Ix1 b) => (a -> b -> c) -> Vector ra a -> Vector rb b -> Vector DS c

Examples

Since: 0.5.0

szipWith3 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c) => (a -> b -> c -> d) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector DS d

Since: 0.5.0

szipWith4 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d) => (a -> b -> c -> d -> e) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector DS e

Since: 0.5.0

szipWith5 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e) => (a -> b -> c -> d -> e -> f) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector DS f

Since: 0.5.0

szipWith6 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f) => (a -> b -> c -> d -> e -> f -> g) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> Vector DS g

Since: 0.5.0

sizipWith :: (Stream ra Ix1 a, Stream rb Ix1 b) => (Ix1 -> a -> b -> c) -> Vector ra a -> Vector rb b -> Vector DS c

Examples

Since: 0.5.0

sizipWith3 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c) => (Ix1 -> a -> b -> c -> d) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector DS d

Since: 0.5.0

sizipWith4 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d) => (Ix1 -> a -> b -> c -> d -> e) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector DS e

Since: 0.5.0

sizipWith5 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e) => (Ix1 -> a -> b -> c -> d -> e -> f) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector DS f

Since: 0.5.0

sizipWith6 :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f) => (Ix1 -> a -> b -> c -> d -> e -> f -> g) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> Vector DS g

Since: 0.5.0

Monadic zipping

szipWithM :: (Stream ra Ix1 a, Stream rb Ix1 b, Monad m) => (a -> b -> m c) -> Vector ra a -> Vector rb b -> m (Vector DS c)

Examples

Since: 0.5.0

szipWith3M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Monad m) => (a -> b -> c -> m d) -> Vector ra a -> Vector rb b -> Vector rc c -> m (Vector DS d)

Since: 0.5.0

szipWith4M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Monad m) => (a -> b -> c -> d -> m e) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> m (Vector DS e)

Since: 0.5.0

szipWith5M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Monad m) => (a -> b -> c -> d -> e -> m f) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> m (Vector DS f)

Since: 0.5.0

szipWith6M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f, Monad m) => (a -> b -> c -> d -> e -> f -> m g) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> m (Vector DS g)

Since: 0.5.0

sizipWithM :: (Stream ra Ix1 a, Stream rb Ix1 b, Monad m) => (Ix1 -> a -> b -> m c) -> Vector ra a -> Vector rb b -> m (Vector DS c)

Examples

Since: 0.5.0

sizipWith3M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Monad m) => (Ix1 -> a -> b -> c -> m d) -> Vector ra a -> Vector rb b -> Vector rc c -> m (Vector DS d)

Since: 0.5.0

sizipWith4M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Monad m) => (Ix1 -> a -> b -> c -> d -> m e) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> m (Vector DS e)

Since: 0.5.0

sizipWith5M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Monad m) => (Ix1 -> a -> b -> c -> d -> e -> m f) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> m (Vector DS f)

Since: 0.5.0

sizipWith6M :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f, Monad m) => (Ix1 -> a -> b -> c -> d -> e -> f -> m g) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> m (Vector DS g)

Examples

Since: 0.5.0

szipWithM_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Monad m) => (a -> b -> m c) -> Vector ra a -> Vector rb b -> m ()

Examples

Since: 0.5.0

szipWith3M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Monad m) => (a -> b -> c -> m d) -> Vector ra a -> Vector rb b -> Vector rc c -> m ()

Since: 0.5.0

szipWith4M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Monad m) => (a -> b -> c -> d -> m e) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> m ()

Since: 0.5.0

szipWith5M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Monad m) => (a -> b -> c -> d -> e -> m f) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> m ()

Since: 0.5.0

szipWith6M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f, Monad m) => (a -> b -> c -> d -> e -> f -> m g) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> m ()

Since: 0.5.0

sizipWithM_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Monad m) => (Ix1 -> a -> b -> m c) -> Vector ra a -> Vector rb b -> m ()

Examples

Since: 0.5.0

sizipWith3M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Monad m) => (Ix1 -> a -> b -> c -> m d) -> Vector ra a -> Vector rb b -> Vector rc c -> m ()

Since: 0.5.0

sizipWith4M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Monad m) => (Ix1 -> a -> b -> c -> d -> m e) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> m ()

Since: 0.5.0

sizipWith5M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Monad m) => (Ix1 -> a -> b -> c -> d -> e -> m f) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> m ()

Since: 0.5.0

sizipWith6M_ :: (Stream ra Ix1 a, Stream rb Ix1 b, Stream rc Ix1 c, Stream rd Ix1 d, Stream re Ix1 e, Stream rf Ix1 f, Monad m) => (Ix1 -> a -> b -> c -> d -> e -> f -> m g) -> Vector ra a -> Vector rb b -> Vector rc c -> Vector rd d -> Vector re e -> Vector rf f -> m ()

Since: 0.5.0

Predicates

Filtering

sfilter :: Stream r ix e => (e -> Bool) -> Array r ix e -> Vector DS e

Sequentially filter out elements from the array according to the supplied predicate.

Example

>>> import Data.Massiv.Array as A
>>> arr = makeArrayR D Seq (Sz2 3 4) fromIx2
>>> arr
Array D Seq (Sz (3 :. 4))
  [ [ (0,0), (0,1), (0,2), (0,3) ]
  , [ (1,0), (1,1), (1,2), (1,3) ]
  , [ (2,0), (2,1), (2,2), (2,3) ]
  ]
>>> sfilter (even . fst) arr
Array DS Seq (Sz1 8)
  [ (0,0), (0,1), (0,2), (0,3), (2,0), (2,1), (2,2), (2,3) ]

Since: 0.5.0

sifilter :: Stream r ix a => (ix -> a -> Bool) -> Array r ix a -> Vector DS a

Similar to sfilter, but map with an index aware function.

Examples

Since: 0.5.0

sfilterM :: (Stream r ix e, Applicative f) => (e -> f Bool) -> Array r ix e -> f (Vector DS e)

Sequentially filter out elements from the array according to the supplied applicative predicate.

Example

>>> import Data.Massiv.Array as A
>>> arr = makeArrayR D Seq (Sz2 3 4) fromIx2
>>> arr
Array D Seq (Sz (3 :. 4))
  [ [ (0,0), (0,1), (0,2), (0,3) ]
  , [ (1,0), (1,1), (1,2), (1,3) ]
  , [ (2,0), (2,1), (2,2), (2,3) ]
  ]
>>> sfilterM (Just . odd . fst) arr
Just (Array DS Seq (Sz1 4)
  [ (1,0), (1,1), (1,2), (1,3) ]
)
>>> sfilterM (\ix@(_, j) -> print ix >> return (even j)) arr
(0,0)
(0,1)
(0,2)
(0,3)
(1,0)
(1,1)
(1,2)
(1,3)
(2,0)
(2,1)
(2,2)
(2,3)
Array DS Seq (Sz1 6)
  [ (0,0), (0,2), (1,0), (1,2), (2,0), (2,2) ]

Since: 0.5.0

sifilterM :: (Stream r ix a, Applicative f) => (ix -> a -> f Bool) -> Array r ix a -> f (Vector DS a)

Similar to filterM, but map with an index aware function.

Corresponds to: filterM (uncurry f) . simap (,)

Examples

Since: 0.5.0

smapMaybe :: Stream r ix a => (a -> Maybe b) -> Array r ix a -> Vector DS b

Apply a function to each element of the array, while discarding Nothing and keeping the Maybe result.

Examples

Since: 0.5.0

smapMaybeM :: (Stream r ix a, Applicative f) => (a -> f (Maybe b)) -> Array r ix a -> f (Vector DS b)

Similar to smapMaybe, but with the Applicative function.

Similar to mapMaybe id $ mapM f arr

Examples

Since: 0.5.0

scatMaybes :: Stream r ix (Maybe a) => Array r ix (Maybe a) -> Vector DS a

Keep all Maybes and discard the Nothings.

Examples

Since: 0.5.0

simapMaybe :: Stream r ix a => (ix -> a -> Maybe b) -> Array r ix a -> Vector DS b

Similar to smapMaybe, but map with an index aware function.

Examples

Since: 0.5.0

simapMaybeM :: (Stream r ix a, Applicative f) => (ix -> a -> f (Maybe b)) -> Array r ix a -> f (Vector DS b)

Similar to smapMaybeM, but map with an index aware function.

Examples

Since: 0.5.0

Folding

sfoldl :: Stream r ix e => (a -> e -> a) -> a -> Array r ix e -> a

Examples

Since: 0.5.0

sfoldlM :: (Stream r ix e, Monad m) => (a -> e -> m a) -> a -> Array r ix e -> m a

Examples

Since: 0.5.0

sfoldlM_ :: (Stream r ix e, Monad m) => (a -> e -> m a) -> a -> Array r ix e -> m ()

Examples

Since: 0.5.0

sifoldl :: Stream r ix e => (a -> ix -> e -> a) -> a -> Array r ix e -> a

Examples

Since: 0.5.0

sifoldlM :: (Stream r ix e, Monad m) => (a -> ix -> e -> m a) -> a -> Array r ix e -> m a

Examples

Since: 0.5.0

sifoldlM_ :: (Stream r ix e, Monad m) => (a -> ix -> e -> m a) -> a -> Array r ix e -> m ()

Examples

Since: 0.5.0

sfoldl1' :: Stream r ix e => (e -> e -> e) -> Array r ix e -> e

Examples

Since: 0.5.0

sfoldl1M :: (Stream r ix e, MonadThrow m) => (e -> e -> m e) -> Array r ix e -> m e

Examples

Since: 0.5.0

sfoldl1M_ :: (Stream r ix e, MonadThrow m) => (e -> e -> m e) -> Array r ix e -> m ()

Examples

Since: 0.5.0

Specialized folds

sor :: Stream r ix Bool => Array r ix Bool -> Bool

Examples

Since: 0.5.0

sand :: Stream r ix Bool => Array r ix Bool -> Bool

Examples

Since: 0.5.0

sall :: Stream r ix e => (e -> Bool) -> Array r ix e -> Bool

Examples

Since: 0.5.0

sany :: Stream r ix e => (e -> Bool) -> Array r ix e -> Bool

Examples

Since: 0.5.0

ssum :: (Num e, Stream r ix e) => Array r ix e -> e

Examples

Since: 0.5.0

sproduct :: (Num e, Stream r ix e) => Array r ix e -> e

Examples

Since: 0.5.0

smaximum' :: (Ord e, Stream r ix e) => Array r ix e -> e

Examples

Since: 0.5.0

smaximumM :: (Ord e, Stream r ix e, MonadThrow m) => Array r ix e -> m e

Examples

Since: 0.5.0

sminimum' :: (Ord e, Stream r ix e) => Array r ix e -> e

Examples

Since: 0.5.0

sminimumM :: (Ord e, Stream r ix e, MonadThrow m) => Array r ix e -> m e

Examples

Since: 0.5.0

Conversions

Lists

stoList :: Stream r ix e => Array r ix e -> [e]

Convert an array to a list by the means of a delayed stream vector.

Related: toList

Examples

Since: 0.5.0

fromList

Arguments

:: Mutable r Ix1 e

=> Comp

Computation startegy to use

-> [e]

Flat list

-> Array r Ix1 e

Convert a flat list into a vector

Since: 0.1.0

sfromList :: [e] -> Vector DS e

Convert a list to a delayed stream vector

Related: fromList, fromListN, sfromListN

Examples

>>> sfromList ([] :: [Int])
Array DS Seq (Sz1 0)
  [  ]
>>> sfromList ([1,2,3] :: [Int])
Array DS Seq (Sz1 3)
  [ 1, 2, 3 ]

Since: 0.5.0

sfromListN :: Sz1 -> [e] -> Vector DS e

Convert a list to a delayed stream vector. Length of the resulting vector will be at most n. This version isn't really more efficient then sfromList, but there is unsafeFromListN

Related: fromList, fromListN, sfromList

Examples

>>> sfromListN 10 [1 :: Int ..]
Array DS Seq (Sz1 10)
  [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 ]
>>> sfromListN 10 [1 :: Int .. 5]
Array DS Seq (Sz1 5)
  [ 1, 2, 3, 4, 5 ]

Since: 0.5.1

Computation

compute :: forall r ix e r'. (Mutable r ix e, Load r' ix e) => Array r' ix e -> Array r ix e

Ensure that Array is computed, i.e. represented with concrete elements in memory, hence is the Mutable type class restriction. Use setComp if you'd like to change computation strategy before calling compute

Since: 0.1.0

computeS :: forall r ix e r'. (Mutable r ix e, Load r' ix e) => Array r' ix e -> Array r ix e

Compute array sequentially disregarding predefined computation strategy. Very much the same as computePrimM, but executed in ST, thus pure.

Since: 0.1.0

computeIO :: forall r ix e r' m. (Mutable r ix e, Load r' ix e, MonadIO m) => Array r' ix e -> m (Array r ix e)

Very similar to compute, but computes an array inside the IO monad. Despite being deterministic and referentially transparent, because this is an IO action it can be very useful for enforcing the order of evaluation. Should be a prefered way of computing an array during benchmarking.

Since: 0.4.5

computePrimM :: forall r ix e r' m. (Mutable r ix e, Load r' ix e, PrimMonad m) => Array r' ix e -> m (Array r ix e)

Compute an array in PrimMonad sequentially disregarding predefined computation strategy.

Since: 0.4.5

computeAs :: (Mutable r ix e, Load r' ix e) => r -> Array r' ix e -> Array r ix e

Just as compute, but let's you supply resulting representation type as an argument.

Examples

>>> import Data.Massiv.Array
>>> computeAs P $ range Seq (Ix1 0) 10
Array P Seq (Sz1 10)
  [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ]

computeProxy :: (Mutable r ix e, Load r' ix e) => proxy r -> Array r' ix e -> Array r ix e

Same as compute and computeAs, but let's you supply resulting representation type as a proxy argument.

Examples

Useful only really for cases when representation constructor or TypeApplications extension aren't desireable for some reason:

>>> import Data.Proxy
>>> import Data.Massiv.Array
>>> computeProxy (Proxy :: Proxy P) $ (^ (2 :: Int)) <$> range Seq (Ix1 0) 10
Array P Seq (Sz1 10)
  [ 0, 1, 4, 9, 16, 25, 36, 49, 64, 81 ]

Since: 0.1.1

computeSource :: forall r ix e r'. (Mutable r ix e, Source r' ix e) => Array r' ix e -> Array r ix e

This is just like convert, but restricted to Source arrays. Will be a noop if resulting type is the same as the input.

Since: 0.1.0

computeWithStride :: forall r ix e r'. (Mutable r ix e, StrideLoad r' ix e) => Stride ix -> Array r' ix e -> Array r ix e

Same as compute, but with Stride.

O(n div k) - Where n is numer of elements in the source array and k is number of elements in the stride.

Since: 0.3.0

computeWithStrideAs :: (Mutable r ix e, StrideLoad r' ix e) => r -> Stride ix -> Array r' ix e -> Array r ix e

Same as computeWithStride, but with ability to specify resulting array representation.

Since: 0.3.0

clone :: Mutable r ix e => Array r ix e -> Array r ix e

O(n) - Make an exact immutable copy of an Array.

Since: 0.1.0

convert :: forall r ix e r'. (Mutable r ix e, Load r' ix e) => Array r' ix e -> Array r ix e

O(n) - conversion between array types. A full copy will occur, unless when the source and result arrays are of the same representation, in which case it is an O(1) operation.

Since: 0.1.0

convertAs :: (Mutable r ix e, Load r' ix e) => r -> Array r' ix e -> Array r ix e

Same as convert, but let's you supply resulting representation type as an argument.

Since: 0.1.0

convertProxy :: (Mutable r ix e, Load r' ix e) => proxy r -> Array r' ix e -> Array r ix e

Same as convert and convertAs, but let's you supply resulting representation type as a proxy argument.

Since: 0.1.1

Deprecated

takeS :: Stream r ix e => Sz1 -> Array r ix e -> Array DS Ix1 e

Deprecated: In favor of stake

See stake.

Since: 0.4.1

dropS :: Stream r ix e => Sz1 -> Array r ix e -> Array DS Ix1 e

Deprecated: In favor of sdrop

See sdrop.

Since: 0.4.1

unfoldr :: (s -> Maybe (e, s)) -> s -> Vector DS e

Deprecated: In favor of sunfoldr

See sunfoldr

Since: 0.4.1

unfoldrN :: Sz1 -> (s -> Maybe (e, s)) -> s -> Vector DS e

Deprecated: In favor of sunfoldrN

See sunfoldrN

Since: 0.4.1

filterS :: Stream r ix e => (e -> Bool) -> Array r ix e -> Array DS Ix1 e

Deprecated: In favor of sfilter

See sfilter

Since: 0.4.1

ifilterS :: Source r ix a => (ix -> a -> Bool) -> Array r ix a -> Array DS Ix1 a

Deprecated: In favor of sifilter

Similar to filterS, but map with an index aware function.

Since: 0.4.1

filterM :: (Stream r ix e, Applicative f) => (e -> f Bool) -> Array r ix e -> f (Vector DS e)

Deprecated: In favor of sfilterM

See sfilterM

Since: 0.4.1

ifilterM :: (Source r ix a, Applicative f) => (ix -> a -> f Bool) -> Array r ix a -> f (Array DS Ix1 a)

Deprecated: In favor of sifilterM

Similar to filterM, but map with an index aware function.

Since: 0.4.1

mapMaybeS :: Stream r ix a => (a -> Maybe b) -> Array r ix a -> Vector DS b

Deprecated: In favor of smapMaybe

See smapMaybe

Since: 0.4.1

imapMaybeS :: Source r ix a => (ix -> a -> Maybe b) -> Array r ix a -> Array DS Ix1 b

Deprecated: In favor of simapMaybe

See simapMaybe

Since: 0.4.1

mapMaybeM :: (Stream r ix a, Applicative f) => (a -> f (Maybe b)) -> Array r ix a -> f (Vector DS b)

Deprecated: In favor of smapMaybeM

See smapMaybeM

Since: 0.4.1

imapMaybeM :: (Source r ix a, Applicative f) => (ix -> a -> f (Maybe b)) -> Array r ix a -> f (Array DS Ix1 b)

Deprecated: In favor of simapMaybeM

See simapMaybeM

Since: 0.4.1

catMaybesS :: Stream r ix (Maybe a) => Array r ix (Maybe a) -> Vector DS a

Deprecated: In favor of scatMaybes

See scatMaybes

Since: 0.4.4

traverseS :: (Stream r ix a, Applicative f) => (a -> f b) -> Array r ix a -> f (Vector DS b)

Deprecated: In favor of straverse

See traverseS

Since: 0.4.5

Re-exports

module Data.Massiv.Core

module Data.Massiv.Array.Delayed

module Data.Massiv.Array.Manifest

module Data.Massiv.Array.Mutable