Vectra: Vector Transformations and Thermostats for ASE Atoms and Molecules

Vectra is a Python package designed to provide tools for vector transformations on ASE atoms and molecules, as well as specialized thermostats for molecular dynamics simulations.

Features

• Decomposition of Quantities: Decompose total quantities (e.g., forces, momenta) into translational, rotational, and vibrational components for individual molecules or groups of atoms.
• Decomposed Berendsen Thermostat: Apply Berendsen thermostats that operate independently on translational, rotational, and vibrational degrees of freedom.

Installation

pip install vectra

Usage

Decomposing Quantities

Decompose a quantity (like forces or velocities) for a set of atoms into translational, rotational, and vibrational components for each identified molecule.

import ase
import numpy as np
from vectra import decompose
from molify import smiles2atoms
from mace.calculators import mace_mp

atoms = smiles2atoms(smiles="O")
atoms.calc = mace_mp() # Any ASE calculator. We are using a uMLIP here.

# Get forces from the calculator
forces = atoms.get_forces()

# Decompose forces
# The 'decompose' function automatically identifies molecules based on connectivity.
decomposed_molecules = decompose(atoms, quantity=forces)

for i, (trans_c, rot_c, vib_c, component_indices) in enumerate(decomposed_molecules):
    print(f"\nMolecule {i+1} (atoms {component_indices}):")
    print("  Translational component:", trans_c)
    print("  Rotational component:", rot_c)
    print("  Vibrational component:", vib_c)

Using Decomposed Berendsen Thermostat

Apply a Berendsen thermostat that can control translational, rotational, and vibrational temperatures independently.

import ase.units
from vectra import DecomposedBerendsenThermostat
from molify import smiles2conformers, pack
from mace.calculators import mace_mp

# Create a system with multiple water molecules using smiles2conformers and pack
water = smiles2conformers(smiles="O", num_conformers=1)[0] # Get one conformer
atoms = pack([water], counts=[3], density=1000)

atoms.calc = mace_mp()

md = DecomposedBerendsenThermostat(
    atoms=atoms,
    timestep=0.5, # fs
    temperature_trans=500,
    temperature_rot=400,
    temperature_vib=300,
    tau_trans=0.5 * 100 * ase.units.fs,
    tau_rot=0.5 * 100 * ase.units.fs,
    tau_vib=0.5 * 100 * ase.units.fs,
)

md.run(1000)