ACEHAL: Hyperactive Learning (HAL) Python interface for building Atomic Cluster Expansion potentials (ACE1.jl/ACE1x.jl)

This package builds ACE interatomic potentials using Hyperactive Learning (HAL). Written by Cas van der Oord and Noam Bernstein.

HAL installation:

1. install Julia 1.8.5 and python 3.9.x (with python ase, scikit-learn, matplotlib and numpy installed)
2. run Julia command

using Pkg; pkg"registry add https://github.com/JuliaRegistries/General"; pkg"registry add https://github.com/ACEsuit/ACEregistry"; pkg"add ACE1, ACE1x, ASE, JuLIP"

make sure you have at least ACE1 version = 0.11.4 and ACE1x = 0.0.4. Use Pkg.activate(".") to use a local project and set environment variable JULIA_PROJECT accordingly. A working Project.toml can be found in /tests/julia_assets/Project.toml

3. install julia Python package to set up Python -> Julia connection

python -m pip install julia==0.6.1

python -c "import julia; julia.install()"

4. Install this package by pip install . or python setup.py install after cloning this repo

ACE1 potentials in Python:

After installation of julia Python package (see 3. above) ACE1x potentials (.json) can be used by first installing pyjulip.

git clone https://github.com/casv2/pyjulip.git
cd pyjulip
pip install .

Python ASE calculators are set up using pyjulip.ACE1("filename.json")

Example scripts

Example scripts can be found in the scripts folder.

References:

If using this code please reference

@misc{van2022hyperactive,
  doi = {10.48550/ARXIV.2210.04225},
  url = {https://arxiv.org/abs/2210.04225},
  author = {van der Oord, Cas and Sachs, Matthias and Kov{\'a}cs, D{\'a}vid P{\'e}ter and Ortner, Christoph and Cs{\'a}nyi, G{\'a}bor},
  title = {Hyperactive Learning (HAL) for Data-Driven Interatomic Potentials},
  publisher = {arXiv},
  year = {2022},
}

@article{DUSSON2022110946,
title = {Atomic cluster expansion: Completeness, efficiency and stability},
journal = {Journal of Computational Physics},
volume = {454},
pages = {110946},
year = {2022},
issn = {0021-9991},
doi = {https://doi.org/10.1016/j.jcp.2022.110946},
url = {https://www.sciencedirect.com/science/article/pii/S0021999122000080},
}