Firedrake — The Complete Guide: Free Automated Finite-Element Solution of Equations in Python
Firedrake is an open-source system from Imperial College London for the automated solution of partial differential equations. You write the problem in Python in a language close to the mathematical formulation, and Firedrake generates and runs fast, parallel code.
Firedrake
Imperial College London and contributors
A Python system for automated finite-element solution of equations: you write the mathematics and the code is generated for you.
Get it from the vendorWhat you get
- Direct mathematical formulation — the UFL language, almost as on paper.
- Automatically fast code — optimised code generation and PETSc solvers.
- Adjoint computation — sensitivities and optimisation without manual derivation.
- Wide fields of use — flow, geophysics, oceanography and solid mechanics.
- Parallel runs — from a laptop to supercomputers.
What to know
Firedrake runs on Linux and macOS, and on Windows through WSL. Installation includes heavy numerical libraries, so it takes some time. It is very similar to FEniCS in concept, and both suit researchers and students familiar with the weak formulation.
Related tools
For the similar, well-known system — FEniCS. For a light library that installs in seconds — scikit-fem. For viewing the results — ParaView.
FEniCS
FEniCS Project
Solve partial differential equations by writing the variational form directly.
Get it from the vendor
scikit-fem
Tom Gustafsson and contributors
A lightweight Python library for assembling and solving finite-element problems, with few dependencies and many examples.
Get it from the vendor
ParaView
Kitware / Sandia / Los Alamos
Large-scale simulation post-processing — slices, streamlines and animation.
Get it from the vendor
Python
Python Software Foundation
The language behind most modern engineering computation, and its full ecosystem.
Get it from the vendorThe bottom line
Firedrake gives the power of a research solver with the convenience of Python — an excellent choice for anyone on Linux who knows the mathematics.