Software used in `Physics-driven machine learning models coupling PyTorch and Firedrake'

This record collates DOIs for the software components used in 'Physics-driven machine learning models coupling PyTorch and Firedrake'.

The Firedrake components and dependencies used were:

• COFFEE (A Compiler for Fast Expression Evaluation): 10.5281/zenodo.3903189
• FInAT (a smarter library of finite elements): 10.5281/zenodo.7623007
• fiat (The Finite Element Automated Tabulator): 10.5281/zenodo.7623001
• firedrake (an automated finite element system): 10.5281/zenodo.7623002
• loopy (Transformation-Based Generation of High-Performance CPU/GPU Code): 10.5281/zenodo.6444324
• petsc (Portable, Extensible Toolkit for Scientific Computation): 10.5281/zenodo.7623006
• tsfc (The Two Stage Form Compiler): 10.5281/zenodo.7623005
• ufl (The Unified Form Language): 10.5281/zenodo.7623003
• PyOP2 (Framework for performance-portable parallel computations on unstructured meshes): 10.5281/zenodo.7623004

You can install Firedrake using exactly this set of component versions using:

firedrake-install --doi 10.5281/zenodo.7623009

See firedrakeproject.org/download.html for more information.