This paper introduces CELL, a Python package designed for cluster expansion (CE) methods, particularly focusing on complex alloys. CELL offers a modular approach, handling various substitutional systems (1D, 2D, 3D alloys) within a multi-component and multi-sublattice framework. It incorporates advanced techniques for dataset creation, model selection, and finite-temperature simulations. The package utilizes the Atomic Simulation Environment (ASE) and scikit-learn, providing visualization tools and interfacing with various ab initio and interatomic potential codes. Examples illustrating CELL's capabilities include a Cu-Pt surface alloy with oxygen adsorption, Si-Ge alloy analysis, and an iterative CE approach for a complex clathrate compound.

Santiago Rigamonti, Maria Troppenz, Martin Kuban, Axel Hübner, Claudia Draxl