pyclupan: Cluster Expansion Tools for Alloy and Substitutional Ionic Systems

pyclupan is a Python package for developing substitutional cluster expansion models based on datasets obtained from density functional theory (DFT) calculations. The code also enables the enumeration of nonequivalent substitutional derivative structures, the calculation of correlation (cluster) functions, Monte Carlo simulations, and the evaluation of free energies for multicomponent substitutional systems.

Citation of Pyclupan

"Cluster expansion method for multicomponent systems based on optimal selection of structures for density-functional theory calculations", A. Seko et al., Phys. Rev. B 80, 165122 (2009)

Required Libraries and Python Modules

• python >= 3.10

• numpy != 2.0.*

• scipy

• pyyaml

• setuptools

• scikit-learn

• pypolymlp >= 0.20.3

• spglib

• phonopy

• graphillion < 2.0

[Optional]

• openmp (recommended)

How to Install Pyclupan

• Install from PyPI

conda create -n pyclupan-env
conda activate pyclupan-env
conda install -c conda-forge pypolymlp spglib phonopy
pip install pyclupan

Installation Notes for macOS

• Install from GitHub

git clone https://github.com/sekocha/pyclupan.git
cd pyclupan
conda create -n pyclupan-env
conda activate pyclupan-env
conda install -c conda-forge pypolymlp spglib phonopy
pip install . -vvv

How to Use Pyclupan

Tutorials

• CE Model Estimation for Alloy Systems (Binary Ag-Au and Ternary Cu-Ag-Au)
• CE Model Estimation for Ionic Substitutional Systems (SrCuO3-x)
• CE Model Estimation for Substitutional Systems on Multiple Sublattices (SiC-AlN)
• Special Quasirandom Structure (SQS) Generation Using Simulated Annealing

Command-Line Interface

• Enumeration of Derivative Structures
• Derivative Structure Sampling
• Enumeration of Clusters and Cluster Function Forms
• Cluster Function Calculations
• Cluster Expansion Model Estimation
• Energy Calculations
• Lattice Monte Carlo Simulations