atomman 0.6

Atomistic Manipulation Toolkit

Atomistic Manipulation Toolkit

AtomMan: the Atomistic Manipulation Toolkit is a Python library for creating, representing, manipulating, and analyzing large-scale atomic systems of atoms. The focus of the package is to facilitate the rapid design and development of simulations that are fully documented and easily adaptable to new potentials, configurations, etc.

The code has no requirements that limit which systems it can be used on, i.e. it should work on Linux, Mac and Windows computers.

The latest release can be installed using pip:

pip install atomman

The code and all documentation is hosted on GitHub and can be directly downloaded at: https://github.com/usnistgov/atomman.

Documentation and examples in the form of Jupyter Notebooks for the various components of the code can be found by clicking on the links in the description below.

1. The core of atomman features:

   • Classes for representing an atomic system:

     • Box class defining a general parallelopid with instant conversions to/from lattice vector, unit cell, and LAMMPS-style parameter representations.

     • Atoms class that stores the per-atom positions and properties for a collection of atoms. The underlying structure of the class uses Numpy arrays and views to efficiently store the values and allow for fast calculations.

     • System class that combines an Atoms object and a System object, along with handling periodic boundaries.

   • Full unit conversion capabilities.

2. The lammps module provides tools for setting up, running, and analyzing LAMMPS simulations directly from Python.

   • General LAMMPS functionality, such as creating systems in LAMMPS, writing input files, reading/writing atom and dump files, calling LAMMPS to run, and recieving LAMMPS errors and log file thermo data.

   • Potential class for reading LAMMPS-potential data models. This class and the data models allow for the modular treatment of any LAMMPS implemented interatomic potentials. In other words, it allows for potentials to be easily swapped.

   • atom_data load() and dump() functions allow for reading/writing of LAMMPS atom coordination files. These functions work with any LAMMPS atom_style and units options.

   • atom_dump load() and dump() functions allow for reading/writing of LAMMPS dump coordination files. Additional JSON files are created that define the proper conversion of the dump files to Python, and provide metadata information to the terms (units, complete name, data structure, etc).

3. The defects module the following components:

   • point defect creation. Add vacancies, substitutionals, positional interstitials, and dumbbell interstitials to a system.

   • Stroh method calculation. Solve the anisotropic elasticity solution for a perfectly straight dislocation of any character using the Stroh method. Once solved, the elastic energy coefficient and position-based displacements and stresses can be obtained. Dislocation monopole systems can be constructed using the Stroh displacements.

4. The models module has tools for reading in Systems:

   • crystal reads in a atomic-system data model file and returns a System.

   • cif has two functions for reading CIF crystallographic files.

     • cif_cell() reads a CIF file and returns a System and list of elements.

     • cif_load() reads a CIF file and returns a DataModelDict representation of all of the lines.

5. The convert module has tools for converting Systems to other codes:

   • ase_Atoms: from_ase_Atoms() and to_ase_Atoms() convert between ase.Atoms classes and atomman.System classes. NOTE: ase has requirements that are not compatible with Windows systems.

   • pymatgen_Structure: from_pymatgen_Structure() and to_pymatgen_Structure() convert between pymatgen.Structure classes and atomman.System classes. NOTE: pymatgen has requirements that are not compatible with Windows systems.

6. The tools module collects all of the additional features that don’t have homes anywhere else.

   • ElasticConstants class for interacting and transforming representations of the elastic constant tensor.

   • nlist function for constructing neighbor lists.

   • other tools for system manipulation, calculations, etc.