A dual-pairing summation-by-parts finite difference framework for nonlinear conservation laws

This repository contains the code to generate the figures and tables presented in https://arxiv.org/abs/2411.06629

Running the code

Clone Repository

First clone the repository and switch to the project directory.

git clone https://github.com/Dougal-s/paper-2024-DP-SBP-nonlinear.git
cd paper-2024-DP-SBP-nonlinear

Install Dependencies

The code requires DrWatson.jl to be installed on the global environment. To do this run the command

julia -e 'using Pkg; Pkg.add("DrWatson")'

To install the project dependencies, run the command

julia --project='.' -e 'using Pkg; Pkg.instantiate()'

Execute Scripts

Each file in scripts corresponds to an individual test case and can be executed by running

julia scripts/the_script_name

Generated figures will be placed in the plots directory and data/tables with be placed in the data directory.

As the scripts cycle through a variety of parameter combinations, it is recommended to run them with the --threads auto flag set to multithreaded execution.

System Image Compilation

Due to the number of scripts, it may be desirable to reduce the time to first execution by generating a custom system image using PackageCompiler.jl. This can be done by executing the script compile/compile.sh. Note that on my machine, this took ~5min and used all 16gb of available memory. Once the system image has been generated, scripts can be run using:

julia -Jpath/to/sysimage.so path/to/script.jl

Authors

• Kenneth Duru (University of Texas at El Paso, US; Australian National University, Canberra, Australia)
• Dougal Stewart (University of Melbourne, Australia)
• Nathan Lee (University of New South Wales, Australia)