Designs for new Base array interface primitives
Author SciML
22 Stars
Updated Last
12 Months Ago
Started In
October 2018


Build Status Build status

Julia has only recently reached v1.0 and the AbstractArray interface is still quite new. The purpose of this library is to solidify extensions to the current AbstractArray interface which are put to use in package ecosystems like DifferentialEquations.jl. Since these libraries are live, this package will serve as a staging ground for ideas before they merged into Base Julia. For this reason, no functionality is exported so that way if such functions are added and exported in a future Base Julia there will be no issues with the upgrade.


A trait function for whether x is a mutable or immutable array. Used for dispatching to in-place and out-of-place versions of functions.


Converts an array of structs formulation to a struct of arrays.


A trait function for whether a matrix x is a sparse structured matrix.


A trait function for whether an array x can use setindex!


Determine whether findstructralnz accepts the parameter x


Returns iterators (I,J) of the non-zeros in the structure of the matrix x. The same as the to first two elements of findnz(::SparseMatrixCSC)


A trait function for whether matrix_colors(A) is a fast algorithm or a slow (graphically-based) method.


Returns an array of for the sparsity colors of a matrix type A. Also includes an abstract type ColoringAlgorithm for matrix_colors(A,alg::ColoringAlgorithm) of non-structured matrices.


A trait function for whether scalar indexing is fast on a given array type.


A getindex which is always allowed.


A setindex! which is always allowed.


Return an instance of the LU factorization object with the correct type cheaply.


Return an instance of the LU factorization object with the correct type cheaply.


Is a form of vec which is safe for all values in vector spaces, i.e. if is already a vector, like an AbstractVector or Number, it will return said AbstractVector or Number.


Creates the zero'd matrix version of u. Note that this is unique because similar(u,length(u),length(u)) returns a mutable type, so is not type-matching, while fill(zero(eltype(u)),length(u),length(u)) doesn't match the array type, i.e. you'll get a CPU array from a GPU array. The generic fallback is u .* u' .* false which works on a surprising number of types, but can be broken with weird (recursive) broadcast overloads. For higher order tensors, this returns the matrix linear operator type which acts on the vec of the array.


Restructures the object y into a shape of x, keeping its values intact. For simple objects like an Array, this simply amounts to a reshape. However, for more complex objects such as an ArrayPartition, not all of the structural information is adequately contained in the type for standard tools to work. In these cases, restructure gives a way to convert for example an Array into a matching ArrayPartition.

List of things to add

Array Types to Handle

The following common array types are being understood and tested as part of this development.

  • Array
  • Various versions of sparse arrays
  • SArray
  • MArray
  • FieldVector
  • ArrayPartition
  • VectorOfArray
  • DistributedArrays
  • GPUArrays (CLArrays and CuArrays)
  • AFArrays
  • MultiScaleArrays
  • LabelledArrays

Breaking Release Notes

2.0: Changed the default of ismutable(array::AbstractArray) = true. We previously defaulted to Base.@pure ismutable(array::AbstractArray) = typeof(array).mutable, but there are a lot of cases where this tends to not work out in a way one would expect. For example, if you put a normal array into an immutable struct that adds more information to it, this is considered immutable, even if all of the setindex! methods work (by forwarding to the mutable array). Thus it seems safer to just always assume mutability is standard for an array, and allow arrays to opt-out.