fluid-roms-bayesian-inference

An experimental repo for constructing ROMs for the numerical solution of statistical inference problems in fluid dynamics. Largely based off of the work of [1] and [2].

Dependencies

1. Git.
2. Julia. As of July 2024, the officially recommended way to get Julia is to install juliaup; it is a popular Julia version manager. On Linux and Mac, you can download and install juliaup (which comes with Julia) in one command:

curl -fsSL https://install.julialang.org | sh

On windows:

winget install julia -s msstore

3. The packages required in the Project.toml. To get them, while running the Julia REPL (run julia), enter package management mode by hitting ], then switch to this project's environment by running activate . (assuming you are in the root directory). Hit backspace to exit package management mode.

Installation

git clone https://github.com/abmarnie/fluid-roms-bayesian-inference.git

Running

I recommend you get Visual Studio Code, as well as the official Julia extension.

References

[1] Jeff Borggaard, Nathan Glatt-Holtz, and Justin Krometis. "A Bayesian approach to estimating background flows from a passive scalar". In: SIAM/ASA Journal on Uncertainty Quantification 8.3 (2020), pp. 1036–1060.

@article{borggaard2020bayesian,
  title={A Bayesian approach to estimating background flows from a passive scalar},
  author={Borggaard, Jeff and Glatt-Holtz, Nathan and Krometis, Justin},
  journal={SIAM/ASA Journal on Uncertainty Quantification},
  volume={8},
  number={3},
  pages={1036--1060},
  year={2020},
  publisher={SIAM}
}

[2] Jeff Borggaard, Nathan E Glatt-Holtz, and Justin Krometis. "A statistical framework for domain shape estimation in Stokes flows". In: Inverse Problems 39.8 (2023), p. 085009.

@article{borggaard2023statistical,
  title={A statistical framework for domain shape estimation in Stokes flows},
  author={Borggaard, Jeff and Glatt-Holtz, Nathan E and Krometis, Justin},
  journal={Inverse Problems},
  volume={39},
  number={8},
  pages={085009},
  year={2023},
  publisher={IOP Publishing}
}

Acknowledgments

This project uses the julia-actions/julia-format GitHub Action, which is under the MIT License.

Some implementation details in this project are largely based off of BayesianShape, InfDimMCM.jl, and FEMfunctions.jl, which are all under the MIT License.

This Julia guide has been very valuable.