rust-dftb 0.1.0

Rust/Rayon implementation for DFTB calculations.

Rust_DFTB

rust_dftb is a Rust/Rayon implementation for DFTB-style calculations with a Python interface. The repository is intended for development and verification of the Rust backend, Python API, and ASE wrapper.

This package does not bundle Slater-Koster parameter files. Pass an external directory containing .skf files through skf_dir.

Requirements

• Python >= 3.10
• Rust toolchain >= 1.76, including cargo
• maturin >= 1.5,<2
• NumPy >= 1.23
• A BLAS/LAPACK backend
• Optional: ASE, for rust_dftb.ase and the ASE examples

The default Cargo feature set uses openblas-system, so the build expects system OpenBLAS with CBLAS/LAPACKE headers and libraries. On Debian/Ubuntu-like systems this usually means installing the OpenBLAS development package before building.

Development install

python -m pip install -U pip maturin numpy
maturin develop --release --features python

For ASE support:

python -m pip install ase

For tests and development tools:

python -m pip install -e ".[dev]"
python -m pytest
cargo test

Minimal Python API example

import numpy as np
from rust_dftb import DftbCalculator, backend_info

z = np.array([1, 1], dtype=np.int64)
positions = np.array(
    [[0.0, 0.0, 0.0], [0.74, 0.0, 0.0]],
    dtype=np.float64,
)

print(backend_info())

calc = DftbCalculator(skf_dir="/path/to/skf", order=2, parallel=True)
print(calc.energy(z, positions))
print(calc.forces(z, positions))

For periodic calculations, pass a cell in Angstrom:

cell = np.eye(3) * 20.0
print(calc.energy(z, positions, cell_angstrom=cell))
print(calc.stress(z, positions, cell_angstrom=cell))

Minimal ASE example

from ase import Atoms
from rust_dftb.ase import ase_calculator

atoms = Atoms("H2", positions=[[0.0, 0.0, 0.0], [0.74, 0.0, 0.0]])
atoms.calc = ase_calculator(skf_dir="/path/to/skf", order=2, parallel=True)

print(atoms.get_potential_energy())
print(atoms.get_forces())

ASE-facing quantities use ASE conventions: positions and cells in Angstrom, energies in eV, forces in eV/Angstrom, and stress in eV/Angstrom^3. The lower-level Python API uses the native units documented in the implementation and examples.

Examples

More complete usage examples are in examples/:

python examples/python_api_all_levels.py --skf-dir /path/to/skf --structure water
python examples/ase_all_levels.py --skf-dir /path/to/skf --structure methane --levels dftb2

See examples/README.md for the longer walkthroughs.

References

The implementation and examples in this repository are related to the following DFTB, SCC-DFTB, DFTB3, multipole-extended DFTB, and ASE references.

1. D. Porezag, T. Frauenheim, T. Köhler, G. Seifert, and R. Kaschner, “Construction of tight-binding-like potentials on the basis of density-functional theory: Application to carbon,” Phys. Rev. B 51, 12947 (1995). DOI: 10.1103/PhysRevB.51.12947.

2. M. Elstner, D. Porezag, G. Jungnickel, J. Elsner, M. Haugk, T. Frauenheim, S. Suhai, and G. Seifert, “Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties,” Phys. Rev. B 58, 7260 (1998). DOI: 10.1103/PhysRevB.58.7260.

3. M. Gaus, Q. Cui, and M. Elstner, “DFTB3: Extension of the self-consistent-charge density-functional tight-binding method (SCC-DFTB),” J. Chem. Theory Comput. 7, 931-948 (2011). DOI: 10.1021/ct100684s.

4. V.-Q. Vuong, T. W. Williams, J. Li, Q. Cui, and I. S. Ufimtsev, “Multipole expansion of atomic electron density fluctuation interactions in the density-functional tight-binding method,” J. Chem. Theory Comput. 19, 8152-8171 (2023). DOI: 10.1021/acs.jctc.3c00778.

5. A. H. Larsen et al., “The atomic simulation environment — a Python library for working with atoms,” J. Phys.: Condens. Matter 29, 273002 (2017). DOI: 10.1088/1361-648X/aa680e.

License

GPL-3.0-or-later.