Overview#

pressio-demoapps is a collection of 1D, 2D and 3D problems of varying complexity (from linear advection, to reaction-diffusion and compressible Euler) that can be used for a variety of purposes.

Is this another physics solver?#

No! This will not become a full solver. The scope is and will stay limited.

Important

The main distinguishing feature from other codes is the built-in support to handle a sample mesh (which, in brief, is the capability of evaluating the residual at a disjoint subset of the mesh cells), and the focus on providing self-contained and well-defined problems to study.

This code, in fact, was originally started as part of the Pressio project to create a suite of benchmark problems to test ROMs and hyper-reduction techniques, but it is being developed to be self-contained. For example, you can just use it for doing “standard” simulations, or you can just use the Python meshing scripts, or leverage the sample mesh capability to study function approximations. One of the objectives is to provide a testbed to explore the impact of varying numerical schemes, and test new sample mesh and hyper-reduction techniques.

Some features of this work include:

  • support for both C++ and Python

  • cell-centered finite volume discretization with various numerical schemes and exact Jacobians

  • built-in support for sample mesh in 1D, 2D and 3D for varying stencil sizes

  • focus on providing self-contained and well-defined problems

Enough text, show me some code!#

A C++ representative snippet showing our idea of a problem:

#include <pressiodemoapps/euler1d.hpp>

int main(){
  namespace pda        = pressiodemoapps;
  const auto meshObj   = pda::load_cellcentered_uniform_mesh_eigen("<path-to-mesh-files>");
  const auto scheme    = pda::InviscidFluxReconstruction::Weno5;
  const auto problemId = pda::Euler1d::Sod;
  auto problem         = pda::create_problem_eigen(meshObj, problemId, scheme);
}

Read more on how to use a problem instance or checkout the C++ public API.

If you prefer, you can use directly Python:

import pressiodemoapps as pda

mesh      = pda.load_cellcentered_uniform_mesh("<path-to-mesh-files>")
scheme    = pda.InviscidFluxReconstruction.Weno5
problemId = pda.Euler1d.Sod
problem   = pda.create_problem(mesh, problemId, scheme)

Read more on how to use a problem instance or checkout the Python public API.

Core development principles#

  • simplicity: we use high-level abstractions and well-defined APIs aiming to make this code as simple as possible to use, thus minimizing the number of steps needed to set up and run a problem. The main idea is a three-step process: (1) generate the mesh, (2) load the mesh and (3) create a problem instance that has a specific API.

  • quality assurance: we maintain a test suite for both the C++ code and Python bindings to ensure stability and reliability;

  • extensibility: easily add new problems, other numerical schemes (e.g. fluxes, reconstruction methods) and different data types.