Metadata-Version: 2.4
Name: mlippykit
Version: 0.0.1
Summary: Pythonic construction of machine-learning potentials
Author-email: Max Hodapp <maxludwig.hodapp@mcl.at>
Project-URL: homepage, https://gitlab.com/mhodapp/mlippykit
Requires-Python: >=3.7
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: ase>=3.24.0
Requires-Dist: atomman>=1.4.9
Requires-Dist: matscipy>=1.1.0
Requires-Dist: numpy>=1.26.4
Requires-Dist: scipy>=1.14.1
Dynamic: license-file

# Introduction

This package contains helper functions for constructing and working with machine-learning interatomic potentials.


# Installation

```
pip install git+https://gitlab.com/mhodapp/mlippykit.git
```


# Description

A short description of the contents of each folder is given below:

- `main folder` - contains some configuration sampling algorithms for constructing training sets for random alloys that have been developed in [Hodapp & Shapeev (2021)](https://doi.org/10.1103/PhysRevMaterials.5.113802)
- `ase_calculators` - wrappers for [ASE](https://wiki.fysik.dtu.dk/ase) calculators
- `configurations` - functionalities for constructing bulk configurations and configurations with defects
- `mlip_utils` - some convenience functions for using MLIP-2 and MLIP-4 functionalities through Python
- `properties` - functions for computing properties like lattice constants, elastic constants, or defect energies
