iLED

This package provides an implementation of the iled modeling approach.

Menier, E., Kaltenbach, S., Yagoubi, M., Schoenauer, M., & Koumoutsakos, P. (2023). Interpretable Learning of Effective Dynamics for multiscale systems, arXiv.

Installation

The package can be installed by first cloning the repository:

git clone git@github.com:emenier/iled.git

Then installing through the following command:

pip3 install -U .

or pip3 install -e . if you intend to contribute.

Training a model for the FHN case

A model for the Fitz-Hugh Nagomo case can be trained as follows:

python examples/FHN/trainscript.py --work_dir /path/to/work/dir --run_name my_run

If unavailable in the specified work_dir, a data directory will be created, in which the data will be automatically downloaded.

Model and training options can be modified directly in the trainscript.py file.

After training, result plots can be generated with:

python examples/FHN/fhn_plotting.py --work_dir /path/to/work/dir --run_name my_run --output examples/FHN/output/

The images will be saved in the specified output directory.