Install.rst

Install

Method 1

Install with pip :

pip install featurebox

Note

If VC++ needed for windows, Please download the dependent packages from Python Extension Packages and install offline. Such as Spglib. And try again, or reference to Method 2.

Method 2

Requirements Packages:

Dependence	Name	Version
necessary	sympy	>=1.6
necessary	deap	>=1.3.1
necessary	scikit-learn	>=0.22.1
necessary	torch	>=1.5.0
necessary	ase
necessary	pymatgen
recommend	scikit-image
recommend	minepy
recommend	torch_geometric

Install by step:

1. sympy :

   pip install sympy>=1.6

Reference: https://www.sympy.org/en/index.html

2. deap :

   pip install deap

Reference: https://github.com/DEAP/deap

3. pymatgen :

   conda install --channel conda-forge pymatgen

Reference: https://github.com/materialsproject/pymatgen

Note

If Spblib needed, which need C++ to compiled, please download Python Extension Packages and pip install locally.

Such as :

pip install /your/local/path/spglib-1.16.1-cp38-cp38-win_amd64.whl

4. scikit-learn :

   conda install sklearn

Reference: https://github.com/materialsproject/pymatgen

5. mgetool:

   pip install mgetool

Reference: https://github.com/Mgedata/mgetool

6. featurebox:

   pip install featurebox

7. ase:

   pip install ase

Reference: https://wiki.fysik.dtu.dk/ase/ , not necessary, just for network.