Installation

minelastic requires Python 3.9 or newer. Its dependencies, numpy, spglib and ASE, are installed with it:

pip install minelastic

The command-line tool is called minelastic. To check the installation:

$ minelastic --version
0.1.0

minelastic --help lists the commands, and minelastic generate --help the options of one of them.

From a clone

The repository also contains the example scripts, the input templates of the tutorials and the tests:

git clone https://github.com/utksi/minelastic.git
cd minelastic
pip install -e ".[test]"
pytest tests

The tests take about a minute and need no DFT code: they use stored stresses of real calculations.

Optional packages

The DFT codes are not needed to install minelastic, to write the inputs or to read the results; they are needed only to run the calculations. Machine-learned potentials are used through their ASE calculators, for example pip install mace-torch for MACE. The tutorial with ASE's EMT potential relaxes the cell with ase.filters.FrechetCellFilter, which needs ASE 3.23 or newer. The documentation is built with pip install -e ".[docs]" followed by mkdocs serve.