Skip to content

maxscheurer/cppe

Repository files navigation

CI Documentation Status Language grade: C/C++

Conda PyPI

CPPE is an open-source, light-weight C++ and Python library for Polarizable Embedding (PE)1,2 calculations. It provides an easy-to-use API to implement PE for ground-state self-consistent field (SCF) calculations and post-SCF methods. A convenient Python interface is also available.

CPPE enables PE calculations in the following programs:

Linear scaling electric field computations in CPPE are achieved through autogenerated code by the fmmgen library.3

Examples for the open-source Python-driven programs can be found here.

Installation

Conda

The easiest way to install CPPE is via conda:

conda install cppe -c conda-forge

Build from Source

Manual builds can be done using CMake by running

git clone https://github.com/maxscheurer/cppe
cd cppe; mkdir build; cd build
cmake ..
make

pip/setuptools

Another way to install CPPE is via pip:

pip install cppe

Note that CPPE will be built from source and a C++14 compatible compiler is required (see below), and OpenMP parallelization is disabled in the setup.py/pip installation. Alternatively, CPPE can be built from source using the setup.py script with

git clone https://github.com/maxscheurer/cppe
cd cppe
python setup.py install

Python interface

If the Python interface should be built, specify the CMake option -DENABLE_PYTHON_INTERFACE=ON. If pybind11 is not installed, CMake will automatically download pybind11 and install it locally. Installing through setup.py will always build the Python interface.

Dependencies

  • C++ 14 compiler
  • Python >= 3.6 (interpreter and development packages)

Tests

The tests can be run with

python setup.py build_ext -i; python setup.py test

for the setup.py build, or

source setup_environment.sh; py.test

for the CMake build.

Citation

Papers:
Code: DOI

CPPE: An Open-Source C++ and Python Library for Polarizable Embedding
Maximilian Scheurer, Peter Reinholdt, Erik Rosendahl Kjellgren, Jógvan Magnus Haugaard Olsen, Andreas Dreuw, and Jacob Kongsted; Journal of Chemical Theory and Computation 2019 15 (11), 6154-6163, DOI: 10.1021/acs.jctc.9b00758

If you use the linear-scaling FMM implementation, please also cite:

Efficient Open-Source Implementations of Linear-Scaling Polarizable Embedding: Use Octrees to Save the Trees
Maximilian Scheurer, Peter Reinholdt, Jógvan Magnus Haugaard Olsen, Andreas Dreuw, and Jacob Kongsted; Journal of Chemical Theory and Computation 2021, DOI: 10.1021/acs.jctc.1c00225

Literature

1 Olsen, J. M. H.; Aidas, K.; Kongsted, J. (2010). Excited States in Solution through Polarizable Embedding. J. Chem. Theory Comput., 6 (12), 3721–3734. https://doi.org/10.1021/ct1003803

2 Olsen, J. M. H.; Kongsted, J. (2011). Molecular Properties through Polarizable Embedding. Advances in Quantum Chemistry (Vol. 61). https://doi.org/10.1016/B978-0-12-386013-2.00003-6

3 Pepper, R.; Fangohr, H. (2020). fmmgen: Automatic Code Generation of Operators for Cartesian Fast Multipole and Barnes-Hut Methods. arXiv:2005.12351