DRDMannTurb (short for Deep Rapid Distortion Theory Mann Turbulence model) is a data-driven framework for synthetic turbulence generation in Python. The code is based on the original work of Jacob Mann in 1994 and 1998 as well as in the deep-learning enhancement developed by Keith et al. in this 2021 publication.
Pre-compiled wheels for the package are available via pip install drdmannturb.
See our development environment instructions
for instructions on installing development versions.
See the /examples/ folder for baselines from the paper and for examples of the many functionalities of the package. These examples are rendered in a more readable
format on our documentation here also.
DRDMannTurb consists of two primary submodules spectra_fitting and fluctuation_generation
which are respectively focused on fitting a Deep Rapid Distortion (DRD) model and
on generating synthetic turbulence "boxes" with a fit DRD model.
If you have any questions, the best way to receive help is by creating a thread in our Discussions or by contacting the authors (Alexey Izmailov, Matthew Meeker) by email directly. If your question pertains to a problem with the package, please open an Issue so that it can addressed.
If you use this software, please cite it as below.
@software{Izmailov_DRDMannTurb_2023,
author = {Izmailov, Alexey and Meeker, Matthew and Deskos, Georgios and Keith, Brendan},
month = mar,
title = {{DRDMannTurb}},
url= {https://github.com/METHODS-Group/DRDMannTurb},
version = {1.0.2},
year = {2024}
}
We always welcome new contributors! The best way to contribute to DRDMannTurb is through opening an issue, making a feature request, or creating a pull request directly.
See also the below instructions for installing DRDMannTurb for development purposes.
DRDMannTurb source is provided as a locally pip-installable package, meaning that
the package can be installed locally with pip install -e . from the package root folder.
In conda_env_files, we provide a conda-lock file,
which supports (64-bit) Intel macOS, ARM macOS, Linux, and Windows,
as well as an environment.yml.
These include all requirements for DRDMannTurb and building the documentation locally,
which is discussed below.
Either of these can be used to create a Conda environment for development
and documentation purposes,
however this is not mandatory.
Recreating the environment off the conda-lock can be done with
conda-lock install -n ENV_NAME. Once this is completed, you should cd .. to
the project root folder and run pip install -e . to install an editable version
of DRDMannTurb in this environment.
If you choose to use your own environment, we ask that you use Python 3.9. You may wish
to reference environment.yml to ensure that all dependencies are installed.
Warning
Due to current incompatibilities between dependencies and Numpy's API
changes in version 2.0, we have currently pinned numpy=1.26.4
as a temporary fix.
We also ask that you install our
pre-commit configuration by running pre-commit install in the root directory
of this repository. If you are unfamiliar with pre-commit,
the documentation can be found here.
DRDMannTurb's test suite is built with Pytest. Running the tests locally can be done by running pytest
from the project root.
Tests decorated with slow can be run with the --runslow flag; they are otherwise skipped. Note that several of these tests require (at least
``partially'') training a DRD model, and so the suite may take several minutes to complete.
Note also that certain components of the test suite require CUDA; these are also
skipped if a CUDA device is not available.
Our documentation source lives in the /docs/ folder.
You should ensure that the dependencies listed in ./requirements-docs.txt are installed;
if you chose to use one of our provided environments, this will have been done already.
Running make html will generate html pages in the /docs/build/html folder; these can be hosted locally with python -m http.server <PORT-NUMBER>.
This work was authored in part by the National Renewable Energy Laboratory, operated by Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. Funding provided by the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy Wind Energy Technologies Office. The views expressed in the article do not necessarily represent the views of the DOE or the U.S. Government. The U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this work, or allow others to do so, for U.S. Government purposes. BK was supported in part by the U.S. Department of Energy Office of Science, Early Career Research Program under Award Number DE-SC0024335.