Python Implementation of Dynamic Mode Decomposition

This Python package provides an implementation of Dynamic Mode Decomposition (DMD) specifically designed for univariate time series forecasting. DMD offers a data-driven approach for analyzing the dynamics of complex systems, making it suitable for forecasting and system identification applications.

Table of Contents

• Python Implementation of Dynamic Mode Decomposition
  • Install dmd
  • Run Examples
  • Uninstall dmd
  • History

Two example scripts are included to demonstrate typical usage scenarios.

For an in-depth explanation of DMD, please refer to the following HackMD note: Dynamic Mode Decomposition.

The primary reference for this implementation is the arXiv article: On Dynamic Mode Decomposition: Theory and Applications.

---

Install dmd

Install from GitHub

pip install git+https://github.com/mukappalambda/dmd.git

Install from Source

git clone https://github.com/mukappalambda/dmd.git
uv build
pip install --user dist/dmd-<version>-py3-none-any.whl

---

Run Examples

cd examples
python dmd_example01.py
python dmd_example02.py

---

Uninstall dmd

pip uninstall dmd -y

History

• Tweak the HackMD note more readable and add the Python code inside that note.