Installation

Vayesta can be installed using pip or from source.

Installing with pip

The simplest way to install Vayesta is to use the setup.py:

[~]$ git clone https://github.com/BoothGroup/Vayesta
[~]$ cd Vayesta
[~]$ pip install .

All required python packages, such as NumPy and PySCF will be installed automatically.

Optional dependencies are required to perform DMET calculations, to leverage MPI parallelism, or to use [ebcc](https://github.com/BoothGroup/ebcc) solvers. These can be specified by adding the modifiers [dmet], [mpi], and [ebcc], respectively.

Installation from Source

Note

Vayesta requires the following python packages when installing from source: NumPy, SciPy, PySCF, and h5py)

To install Vayesta from source, clone the GitHub repository and use cmake and make to compile:

[~]$ git clone https://github.com/BoothGroup/Vayesta .
[~]$ cd Vayesta/vayesta/libs
[~]$ mkdir build && cd build
[~]$ cmake ..
[~]$ make

To ensure that Vayesta is found by the Python interpreter when calling import vayesta, the installation location needs to be prepended to the PYTHONPATH environment variable as follows:

[~]$ export PYTHONPATH:"/path/to/vayesta":$PYTHONPATH

This command can be added to the ~/.profile file located in the home dirctory, to ensure that it is excecuted for every new shell instance.

Running Tests

After installation it is a good idea to run the test suite with pytest using the following command:

[~]$ pytest vayesta/tests