MeshArrays.jl

MeshArrays.jl primarily defines composite types that embed inter-connected array collections and provides exchange functions that transfer data between connected arrays (see Notebooks below). It was originally introduced in this JuliaCon-2018 presentation. Note: MeshArrays.jl is registered, documented, etc., but still regarded as a preliminary implementation.

Installation

using Pkg
Pkg.add("MeshArrays")
Pkg.test("MeshArrays")

Use example

Let's integrate a diffusion equation over the surface of a cube:

using MeshArrays; p=dirname(pathof(MeshArrays));

GridVariables=GridOfOnes("cs",6,100)
DemoVariables=MeshArrays.demo2(GridVariables)

using Plots; include(joinpath(p,"Plots.jl"));
heatmap(DemoVariables[2],clims=(-0.5,0.5))

Starting from a noisy DemoVariables[1], this leads to a smoothed DemoVariables[2].

Notebooks

The JuliaCon-2018 presentation relied on two Jupyter notebooks that are available in the MeshArrayNotebooks repo. Other included notebooks illustrate for example how MeshArrays.smooth is used for unit testing purposes (demo_smooth.ipynb) or how MeshArrays.jl allows for computations of Ocean transports (demo_trsp.ipynb). All MeshArrayNotebooks can readily be used in the cloud via binder.