Skip to content

MPI-supported Scalable One-stop Platform for Hydroelastic Things (SOPHT)

License

Notifications You must be signed in to change notification settings

SophT-Team/sopht-mpi

 
 

Repository files navigation

SophT-MPI

CI DOI

Scalable One-stop Platform for Hydroelastic Things (SOPHT) MPI solver.

Python implementation of an elastohydrodynamic MPI solver, for resolving flow-structure interaction of 3D mixed soft/rigid bodies in viscous flows.

Installation

Below are steps of how to install sopht-mpi. We mainly use poetry to manage the project, although most of the important commands will be provided in Makefile.

  1. Clone!

First create the fork repository and clone to your local machine.

  1. Virtual python workspace: conda.

We recommend using python version above 3.10.

conda create --name sopht-mpi-env
conda activate sopht-mpi-env
conda install python==3.10
  1. Install non-python dependencies, that include MPI, hdf5-mpi and fftw. For Ubuntu you can use:
make install_non_python_modules_on_ubuntu

And for Mac-OS one can use

make install_non_python_modules_on_macos

NOTE: For cluster, optimised versions of the above modules are already installed, and as such this step can be skipped, and the internal modules should be loaded directly.

  1. Setup poetry and dependencies!
make poetry-download
make install
make pre-commit-install

Citation

We ask that any publications which use SophT-MPI cite as following:

@software{fan_kiat_chan_2023_7659153,
  author       = {Fan Kiat Chan and
                  Yashraj Bhosale},
  title        = {{Scalable One-stop Platform for Hydroelastic Things
                   (SOPHT) MPI solver}},
  month        = feb,
  year         = 2023,
  publisher    = {Zenodo},
  version      = {0.0.1},
  doi          = {10.5281/zenodo.7659153},
  url          = {https://doi.org/10.5281/zenodo.7659153}
}

About

MPI-supported Scalable One-stop Platform for Hydroelastic Things (SOPHT)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 99.6%
  • Makefile 0.4%