ChemPotPy, CHEMical library of POTential energy surfaces in PYthon

January 17, 2025 ยท View on GitHub

ChemPotPy, CHEMical library of POTential energy surfaces in PYthon

Jan. 17, 2025

Authors: Yinan Shu, Zoltan Varga, Dayou Zhang, Donald G. Truhlar University of Minnesota, Minnesota, United States

ChemPotPy is a library for analytic representation of single-state and multi-state potential energy surfaces and couplings.

All fortran source code are stored in folder chempotpy

How to install

  • Create a conda virtual environment with gfortran and MKL:

     conda create --name chempotpy
 conda activate chempotpy
 conda install python=3.11
 conda install mkl mkl-service
 conda install -c conda-forge gfortran
 pip install numpy "numpy>=1.26,<1.27"
 pip install charset_normalizer

  • Install stable release:

      pip install chempotpy

  • The users may re-compile all .so modules for compatability reasons:

    get into the parent directory of chempotpy

      make all 
  make check

Compile Chempotpy subroutine

One can use the meta programming script to automatically generate a fortran subroutine.

create a conda vitural environment as suggested in How to install

get into the parent directory of chempotpy/chempotpy

   ./meta_chempotpy.script

The meta program will generate a fortran subroutine called chempotpy. One can interface this chempotpy subroutine with any dynamics code. Notice that the meta program will also generate sub programs for each surface that is located in chempotpy/chempotpy/system/lib/.

Citation

The following paper should be cited in publications utilizing the ChemPotPy library in addition to the original paper that publishes the potential energy surface subroutine:

Shu, Y.; Varga, Z.; Truhlar, D. G. "ChemPotPy: A Python Library for Analytic Representation of Potential Energy Surfaces and Diabatic Potential Energy Matrices" accepted by J. Phys. Chem. A

TO CONTRIBUTE YOUR POTENTIAL or report a bug