RecursiveArrayTools.jl is a set of tools for dealing with recursive arrays like arrays of arrays. The current functionality includes:
VectorOfArray is an array which has the underlying data structure
(but hopefully concretely typed!). This wrapper over such data structures allows one to lazily
act like it's a higher dimensional vector, and easily convert to different forms. The indexing
A[i] # Returns the ith array in the vector of arrays A[j,i] # Returns the jth component in the ith array A[j1,...,jN,i] # Returns the (j1,...,jN) component of the ith array
which presents itself as a column-major matrix with the columns being the arrays from the vector.
AbstractArray interface is implemented, giving access to
append!, etc. function
which act appropriate. Points to note are:
- The length is the number of vectors, or
uis the vector of arrays.
- Iteration follows the linear index and goes over the vectors
convert(Array,VA::AbstractVectorOfArray) function is provided which transforms
VectorOfArray into a matrix/tensor. Also,
returns a vector of the series for each component, that is
A[i,:] for each
A plot recipe is provided which plots the
Related to the
VectorOfArray is the
This is a
VectorOfArray which stores
A.t which matches
A.u. This will plot
(A.t[i],A[i,:]). The function
tuples(diffeq_arr) returns tuples of
A is an array which is made up of different arrays
These index like a single array, but each subarray may have a different type.
However, broadcast is overloaded to loop in an efficient manner, meaning that
A .+= 2.+B is type-stable in its computations, even if
do not match types. A full array interface is included for completeness, which
allows this array type to be used in place of a standard array in places where
such a type stable broadcast may be needed. One example is in heterogeneous
differential equations for DifferentialEquations.jl.
ArrayPartition acts like a single array.
A[i] indexes through the first
array, then the second, etc. all linearly. But
A.x is where the arrays are stored.
using RecursiveArrayTools A = ArrayPartition(y,z)
We would have
A.x==z. Broadcasting like
f.(A) is efficient.
copy! function. Acts like a
deepcopy! on arrays of arrays, but
copy! on arrays of scalars.
Technically just a Base fallback that works well. Takes in a vector of arrays,
returns an array of dimension one greater than the original elements.
AbstractVectorOfArray. If the
vecvec is ragged, i.e. not all of the
elements are the same, then it uses the size of the first element to determine
f on each element of a vecvec
i<length(x), it's simply a
recursivecopy! to the
ith element. Otherwise it will
one on the bottom container to get the "true element one type"
Generalized mean functions for vectors of arrays and matrix of arrays.