E3x is a JAX library for constructing efficient \mathrm{E}(3)-equivariant deep learning architectures built on top of Flax.
To learn how E3x works, what \mathrm{E}(3)-equivariance means, and for what it is useful, please have a look at the :ref:`Overview`, which also introduces notation used throughout the documentation. To learn how to use E3x, please refer to the :ref:`Examples`, which show how to solve common tasks with simple toy problems. If you encounter any difficulties or problems when working with E3x, make sure to check the :ref:`Pitfalls` for common mistakes and sources of error. More details on the mathematical theory behind E3x can be found in this paper.
E3x is available on github. To install E3x, simply run
python -m pip install --upgrade e3x
.. toctree:: :maxdepth: 1 :caption: Quickstart overview pitfalls
.. toctree:: :maxdepth: 1 :caption: Examples examples/tetracubes examples/moment_of_inertia examples/md17_ethanol
.. toctree:: :maxdepth: 1 :caption: Useful recipes and tricks changing_defaults neighbor_lists basis_functions constructing_cartesian_tensors
.. toctree:: :maxdepth: 3 :caption: API reference api