An Array is a vector with a shape.

Class for types that can be converted to and from linear indexes.

Boxed array.

Unboxed array.

Storeable array.

Prim array.

Class of things that have a Layout. This means we can use the same functions for the various different arrays in the library.

A Layout is the full size of an array. This alias is used to help distinguish between the layout of an array and an index (usually just l Int) in a type signature.

Get the extent of an array.

extent :: Array v f a    -> f Int
extent :: MArray v f s a -> f Int
extent :: Delayed f a    -> f Int
extent :: Focused f a    -> f Int

Get the total number of elements in an array.

size :: Array v f a    -> Int
size :: MArray v f s a -> Int
size :: Delayed f a    -> Int
size :: Focused f a    -> Int

Indexed fold for all the indexes in the layout.

Indexed fold starting starting from some point, where the index is the linear index for the original layout.

Indexed fold between the two indexes where the index is the linear index for the original layout.

Indexed lens over the underlying vector of an array. The index is the extent of the array. You must _not_ change the length of the vector, otherwise an error will be thrown (even for V1 layouts, use flat for V1).

Indexed traversal over the elements of an array. The index is the current position in the array.

Same as values but restrictive in the vector type.

Traverse over the values between two indexes.

1D arrays are just vectors. You are free to change the length of the vector when going over this Iso (unlike linear).

Note that V1 arrays are an instance of Vector so you can use any of the functions in Data.Vector.Generic on them without needing to convert.

Contruct a flat array from a list. (This is just fromList from Generic.)

O(n) Convert the first n elements of a list to an Array with the given shape. Returns Nothing if there are not enough elements in the list.

O(n) Convert the first n elements of a list to an Array with the given shape. Throw an error if the list is not long enough.

Create an array from a vector and a layout. Return Nothing if the vector is not the right shape.

Create an array from a vector and a layout. Throws an error if the vector is not the right shape.

O(n) Array of the given shape with the same value in each position.

O(n) Construct an array of the given shape by applying the function to each index.

O(n) Construct an array of the given shape by applying the function to each index.

Execute the monadic action and freeze the resulting array.

O(n) Construct an array of the given shape by filling each position with the monadic value.

O(n) Construct an array of the given shape by applying the monadic function to each index.

O(n) Construct an array of the given shape by applying the monadic function to each index.

The empty Array with a zero shape.

Test is if the array is empty.

Index an element of an array. Throws IndexOutOfBounds if the index is out of bounds.

Safe index of an element.

Index an element of an array without bounds checking.

Index an element of an array while ignoring its shape.

Index an element of an array while ignoring its shape, without bounds checking.

O(1) Indexing in a monad.

The monad allows operations to be strict in the vector when necessary. Suppose vector copying is implemented like this:

copy mv v = ... write mv i (v ! i) ...

For lazy vectors, v ! i would not be evaluated which means that mv would unnecessarily retain a reference to v in each element written.

With indexM, copying can be implemented like this instead:

copy mv v = ... do
  x <- indexM v i
  write mv i x

Here, no references to v are retained because indexing (but not the elements) is evaluated eagerly.

Throws an error if the index is out of range.

O(1) Indexing in a monad without bounds checks. See indexM for an explanation of why this is useful.

O(1) Indexing in a monad. Throws an error if the index is out of range.

O(1) Indexing in a monad without bounds checks. See indexM for an explanation of why this is useful.

For each pair (i,a) from the list, replace the array element at position i by a.

O(m+n) For each pair (i,b) from the list, replace the array element a at position i by f a b.

O(n) Map a function over an array

O(n) Apply a function to every element of a vector and its index

Zip two arrays element wise. If the array's don't have the same shape, the new array with be the intersection of the two shapes.

Zip three arrays element wise. If the array's don't have the same shape, the new array with be the intersection of the two shapes.

Zip two arrays using the given function. If the array's don't have the same shape, the new array with be the intersection of the two shapes.

Zip three arrays using the given function. If the array's don't have the same shape, the new array with be the intersection of the two shapes.

Zip two arrays using the given function with access to the index. If the array's don't have the same shape, the new array with be the intersection of the two shapes.

Zip two arrays using the given function with access to the index. If the array's don't have the same shape, the new array with be the intersection of the two shapes.

Affine traversal over a single row in a matrix.

>>> traverseOf_ rows print $ m & ixRow 1 . each *~ 2
[a,b,c,d]
[e * 2,f * 2,g * 2,h * 2]
[i,j,k,l]

The row vector should remain the same size to satisfy traversal laws but give reasonable behaviour if the size differs:

>>> traverseOf_ rows print $ m & ixRow 1 .~ B.fromList [0,1]
[a,b,c,d]
[0,1,g,h]
[i,j,k,l]

>>> traverseOf_ rows print $ m & ixRow 1 .~ B.fromList [0..100]
[a,b,c,d]
[0,1,2,3]
[i,j,k,l]

Indexed traversal over the rows of a matrix. Each row is an efficient slice of the original vector.

>>> traverseOf_ rows print m
[a,b,c,d]
[e,f,g,h]
[i,j,k,l]

Affine traversal over a single column in a matrix.

>>> traverseOf_ rows print $ m & ixColumn 2 . each +~ 1
[a,b,c + 1,d]
[e,f,g + 1,h]
[i,j,k + 1,l]

Indexed traversal over the columns of a matrix. Unlike rows, each column is a new separate vector.

>>> traverseOf_ columns print m
[a,e,i]
[b,f,j]
[c,g,k]
[d,h,l]

>>> traverseOf_ rows print $ m & columns . indices odd . each .~ 0
[a,0,c,0]
[e,0,g,0]
[i,0,k,0]

The vectors should be the same size to be a valid traversal. If the vectors are different sizes, the number of rows in the new array will be the length of the smallest vector.

Traversal over a single plane of a 3D array given a lens onto that plane (like _xy, _yz, _zx).

Traversal over all planes of 3D array given a lens onto that plane (like _xy, _yz, _zx).

Flatten a plane by reducing a vector in the third dimension to a single value.

This Traversal should not have any duplicates in the list of indices.

A mutable array with a shape.

Boxed mutable array.

Unboxed mutable array.

Storable mutable array.

Primitive mutable array.

O(n) Yield an immutable copy of the mutable array.

O(n) Yield a mutable copy of the immutable vector.

O(1) Unsafely convert an immutable array to a mutable one without copying. The immutable array may not be used after this operation.

O(1) Unsafe convert a mutable array to an immutable one without copying. The mutable array may not be used after this operation.

A delayed representation of an array. This useful for mapping over an array in parallel.

Isomorphism between an array and its delayed representation. Conversion to the array is done in parallel.

Isomorphism between an array and its delayed representation. Conversion to the array is done in parallel.

Turn a material array into a delayed one with the same shape.

Parallel manifestation of a delayed array into a material one.

Sequential manifestation of a delayed array.

Generate a Delayed array using the given Layout and construction function.

Index a delayed array, returning a IndexOutOfBounds exception if the index is out of range.

manifest an array to a UArray and delay again. See Data.Dense.Boxed or Data.Dense.Storable to affirm for other types of arrays.

seqManifest an array to a UArray and delay again. See Data.Dense.Boxed or Data.Dense.Storable to affirm for other types of arrays.

A delayed representation of an array with a focus on a single element. This element is the target of extract.

Focus on a particular element of a delayed array.

Discard the focus to retrieve the delayed array.

Indexed lens onto the delayed array, indexed at the focus.

Modify a Delayed array by extracting a value from a Focused each point.

Lens onto the position of a ComonadStore.

locale :: Lens' (Focused l a) (l Int)

Focus on a neighbouring element, relative to the current focus.

The boundary condition used for indexing relative elements in a Focused.

Index a focused using a Boundary condition.

Index an element by applying a function the current position, using a boundary condition.

Index an element relative to the current focus using a Boundary condition.

Generate a stream from a Layout's indices.

Generate a stream from a Layout's indices.

Make a stream of the indexes of a Layout.

Generate a bundle from Layout indices.

Generate a bundle from Layout indices.

Generate a bundle of indexes for the given Layout.