pyiron_lammps

The pyiron_lammps packages provides primarily two functions. A write_lammps_structure() function to write an ase.atoms.Atoms structure to an LAMMPS data file and a parse_lammps_output_files() function to parse the log.lammps, dump.out and dump.h5 files from the LAMMPS thermo and dump commands.

Installation

The pyiron_lammps package is distributed via both pypi:

pip install pyiron-lammps

and conda-forge:

conda install -c conda-forge pyiron_lammps

Write LAMMPS structure

The write_lammps_structure() function is designed to write an ase.atoms.Atoms structure to an LAMMPS data file:

from pyiron_lammps import write_lammps_structure

write_lammps_structure(
    structure,
    potential_elements,
    units="metal",
    file_name="lammps.data",
    working_directory=None,
)

The structure parameter refers to the ase.atoms.Atoms structure and the potential_elements refers to the list of elements implemented in the specific interatomic potential. For example the NiAlH_jea.eam.alloy potential implements the elements Ni, Al and H, so when writing a structure for a simulation with this potential the potential_elements=["Ni", "Al", "H"]. It is important to maintain the order of the elements as LAMMPS internally references the elements based on their index, starting from one. The units parameter refers to the LAMMPS internal units to convert the ase.atoms.Atoms object which is defined in Angstrom to the length scale of the LAMMPS simulation. Finally, file_name parameter and the current working directory working_directory parameter are designed to select the location where the LAMMPS structure should be written. With the default parameters the LAMMPS structure is written in the lammps.data file in the current directory.

Parse LAMMPS output

In addition to writing the LAMMPS input structure pyiron_lammps also provide the parse_lammps_output_files() function to parse the LAMMPS output files, namely the log.lammps, dump.out and dump.h5 files:

from pyiron_lammps import parse_lammps_output_files

parse_lammps_output_files(
    working_directory,
    structure,
    potential_elements,
    units="metal",
    dump_h5_file_name="dump.h5",
    dump_out_file_name="dump.out",
    log_lammps_file_name="log.lammps",
)

In analogy to the write_lammps_structure() function the working_directory parameter refers to the directory which contains the output files. The structure parameter reefers to the ase.atoms.Atoms object which should be used as template to parse the structure from the dump files. This structure is again required as LAMMPS internally references elements only by an index, so the template structure is required to map the elements from the interatomic potential back to the elements of the ase.atoms.Atoms object. In the same way the potential_elements refers to the list of elements implemented in the specific interatomic potential. The units parameter refers to the LAMMPS internal units to convert the ase.atoms.Atoms object which is defined in Angstrom to the length scale of the LAMMPS simulation. Finally, the parameters dump_h5_file_name, dump_out_file_name and log_lammps_file_name refer to the output file names.

For the dump.out file the following LAMMPS dump command should be added to the LAMMPS input file:

dump 1 all custom 100 dump.out id type xsu ysu zsu fx fy fz vx vy vz
dump_modify 1 sort id format line "%d %d %20.15g %20.15g %20.15g %20.15g %20.15g %20.15g %20.15g %20.15g %20.15g"

For the log.lammps file the following LAMMPS thermo command should be added to the LAMMPS input file:

thermo_style custom step temp pe etotal pxx pxy pxz pyy pyz pzz vol
thermo_modify format float %20.15g
thermo 100

Usage

Currently, the pyiron_lammps parser is primarily used in the pyiron_atomistics package and its successor the atomistics package to provide a simple LAMMPS parser. It only depends on ase, numpy, pandas and scipy and has an optional dependency on h5py to parse the LAMMPS h5md format.