calorine

calorine is a Python library for constructing and sampling neuroevolution potential (NEP) models via the GPUMD package. It provides ASE calculators, IO functions for reading and writing GPUMD input and output files, as well as a Python interface that allows inspection of NEP models. Tutorials with common usage examples can be found under Tutorials, and a detailed function reference can be found under Function reference.

The following snippet illustrates how a CPUNEP calculator instance can be created given a NEP potential file, and how it can be used to predict the potential energy, forces, and stress for a structure.

from ase.io import read
from ase.build import bulk
from calorine.calculators import CPUNEP

structure = bulk('PbTe', crystalstructure='rocksalt', a=6.7)
calc = CPUNEP('nep-PbTe.txt')
structure.calc = calc

print('Energy (eV):', structure.get_potential_energy())
print('Forces (eV/Å):\n', structure.get_forces())
print('Stress (GPa):\n', structure.get_stress())

If you use calorine in any of your publications, please follow the citation instructions for how to properly cite this work.

calorine and its development are hosted on gitlab.

Main

• Installation
  • Installation via pip
  • Installing the development version
• Credits

Tutorials

• Using the ASE calculators
  • CPU-based calculator
  • GPU-based calculator
• Accessing NEP descriptors
  • Accessing latent space representations
• Inspection of a NEP model
  • Parsing a model file and basic inspection
  • Analysis of the descriptors
  • Manipulating a model
  • Peeling the model, layer by layer
• Descriptor space analysis
  • Setup example system and calculate descriptors
  • Performing PCA on the descriptors
  • Normalized vs non-normalized descriptors
• Structure generation and NEP training
  • Structure generation
  • Training
  • Prototype structures
  • Generation of strained and deformed prototype structures
  • Generation of rattled structures
  • Training
  • Analysis of the model
• Structure relaxation
  • Relaxation with a fixed cell
• Phonon properties
  • Structure preparation
  • Compute force constants
  • Phonon dispersion
  • Density of states
• Elastic properties
  • Face-centered cubic silver
  • CsPbI3
• Free energy calculations
  • Introduction
  • Frenkel-Ladd approach
  • Reversible scaling approach
• Thermal conductivity from BTE
  • Workflow
  • BTE computational parameters
  • Miscellaneous
  • Preparations
  • Relaxation of primitive cell
  • Construction of force constants
  • BTE calculation
  • Analysis of results

Function reference

• ASE calculators
  • CPU calculator
  • GPU calculator
• GPUMD IO
  • read_hac()
  • read_kappa()
  • read_mcmd()
  • read_msd()
  • read_runfile()
  • read_thermo()
  • read_thermodynamic_data()
  • read_xyz()
  • write_runfile()
  • write_xyz()
• NEP interface
  • get_descriptors()
  • get_dipole()
  • get_dipole_gradient()
  • get_latent_space()
  • get_parity_data()
  • get_polarizability()
  • get_polarizability_gradient()
  • get_potential_forces_and_virials()
  • read_loss()
  • read_model()
  • read_nepfile()
  • read_structures()
  • setup_training()
  • write_nepfile()
  • write_structures()
  • Model
• Additional functions
  • analyze_data()
  • get_autocorrelation_function()
  • get_correlation_length()
  • get_elastic_stiffness_tensor()
  • get_error_estimate()
  • get_force_constants()
  • get_primitive_structure()
  • get_rtc_from_hac()
  • get_spacegroup()
  • get_wyckoff_sites()
  • relax_structure()

Backmatter

• Index