This python code illustrates how to apply Dynamic Mode Decomposition (DMD) to univariate time series forecasting tasks.

Two examples are provided here.

For the explanation of DMD, please refer to this HackMD notes: Dynamic Mode Decomposition.

The major reference is this arXiv article: On Dynamic Mode Decomposition: Theory and Applications.

Install from source:

git clone https://github.com/mukappalambda/dmd.git
poetry build
cd dist
pip install dmd-*.whl
#pip show dmd

Run the examples:

cd examples
python dmd_example01.py
python dmd_example02.py

Uninstall:

pip uninstall dmd -y

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