Skip to content

PyQBMMlib is a Python extension of QBMMlib

License

Notifications You must be signed in to change notification settings

comp-physics/PyQBMMlib

Repository files navigation

PyQBMMlib

CI Documentation Status DOI

PyQBMMlib is a Python fork of QBMMlib, which was developed by Prof. Spencer Bryngelson, Prof. Rodney Fox, and Prof. Tim Colonius. It can be cited as

@article{bryngelson_2020,
    Author        = {Spencer H. Bryngelson and Tim Colonius and Rodney O. Fox},
    Title         = {{QBMMlib: A} library of quadrature-based moment methods},
    Journal       = {SoftwareX},
    Volume        = {12},
    Pages         = {100615},
    Year          = {2020},
}

When compared to QBMMlib, PyQBMMlib offers significantly decreased time to solution (when using Numba).

PyQBMMlib is used to train new QBMMs based on neural networks in this paper.

Authors:

Required Python modules

  • Python >= 3.0
  • Numpy
  • Scipy
  • Sympy
  • Optional: Numba (significant speedup via JIT compiling)

Current capabilities

PyQBMMlib is under active development. However, it still has all the capabilities of QBMMlib except for traditional CQMOM, which was elided in lieu of its contemporary, CHyQMOM. This includes:

  • Automatic formulation of moment transport equations
  • 1-3D moment inversion
  • QMOM (Wheeler), HyQMOM (2 and 3 node), CHyQMOM (4, 9, and 27 node)
  • SSP RK2-3

Features under development

Several more features will be added to PyQBMMlib. A partial list is included here.

  • 2D + static 1D moment inversion
  • Spatial dependencies and fluxes (3D flows)

Acknowledgements

Great thanks is owed to Professor Alberto Passalacqua (Iowa State University) for his part in developing these algorithms and teaching me the same. We acknowledge funding from the U.S. Office of Naval Research under grant numbers N0014-17-1-2676 and N0014-18-1-2625 (SHB) and the U.S. Department of Energy, National Nuclear Security Administration, under Award Number DE-NA0002374 (ECG).

License

PyQBMMlib is under the MIT license.