Update

2024/07/23: Add Fe05_zbl.
2024/07/23: Rename pair_aenetcc.* pair_aenet_cc.* and add weight_functions.h.
2024/03/16: Update pair_aenetcc.* and aenet_sfb.h.

Installation

0. This manual explains only clean install case using gnu g++, gfortran, and openmpi under linux OS.

1. Clone this package to your appropriate directory.

git clone https://github.com/HidekiMori-CIT/aenetcc-pre.git

2. Download and unpack LAMMPS package (currently stable_2Aug2023.tar.gz) from Github to same directory of aenetcc-pre:

wget https://github.com/lammps/lammps/archive/refs/tags/stable_2Aug2023.tar.gz
tar -xvzf stable_2Aug2023.tar.gz

3. Copy USER-ML-AENET-CC/ in aenetcc-pre/ to lammps-stable_2Aug2023/src/ and complie LAMMPS with aenet-cc module.

cp -r ./aenetcc-pre/USER-ML-AENET-CC/ ./lammps-stable_2Aug2023/src/  
cd ./lammps-stable_2Aug2023/src/  
make yes-user-ml-aenet-cc
make mpi

note: If you use intel compiler, replace mpi to icc_openmpi. Other compile option of LAMMPS, please see LAMMPS manual.

Citing of this package and ANN potential

[1] H. Mori and T. Ozaki, Phys. Rev. Mater., 4, 040601 (2020).
[2] H. Mori, et. al., Phys. Rev. Matter., 7, 063605 (2023).