A Python package for the application of Reduced Order Modeling (ROM) to large datasets using the Singular Value Decomposition (SVD). The backbone of SVD-ROM is the truncated SVD, which allows you to perform dimensionality reduction on huge arrays, and implement machine learning methods such as Principal Component Analysis (PCA), Proper Orthogonal Decomposition (POD), Spectral Proper Orthogonal Decomposition (sPOD), or Dynamic Mode Decomposition (DMD). These methods have applications in fields such as fluid dynamics, combustion, finance, weather and climate modeling, neuroscience, or chemometrics, to name a few. SVD-ROM is work in progress, and currently supports (or will soon support) PCA, POD and DMD. Other methods will be implemented in the future.
From source:
git clone https://github.com/dsj976/svdrom
cd svdrom
python -m pip install .The best way to get started is to have a look at the notebooks in the demos folder.
See CONTRIBUTING.md for instructions on how to contribute.
Distributed under the terms of the MIT license.
 Robert Vava 👀 💻 🚧 |
 David Salvador Jasin 💻 🚧 👀 🚇 |