Adaptive probabilistic ODE solvers without adaptive memory requirements

This repository contains the experiments for:

@InProceedings{kramer2024adaptive,
  title     = {Adaptive Probabilistic ODE Solvers Without Adaptive Memory Requirements},
  author    = {Kr\"{a}mer, Nicholas},
  booktitle = {Proceedings of the First International Conference on Probabilistic Numerics},
  pages     = {12--24},
  year      = {2025},
  editor    = {Kanagawa, Motonobu and Cockayne, Jon and Gessner, Alexandra and Hennig, Philipp},
  volume    = {271},
  series    = {Proceedings of Machine Learning Research},
  publisher = {PMLR},
  url       = {https://proceedings.mlr.press/v271/kramer25a.html}
}

Here is a link to the arXiv.

Warning

This repository contains only experiment code. We have made the new algorithm a part of probdiffeq (link), which can be installed via

pip install probdiffeq

Unless you are looking for specific experiment configurations, you are probably interested in probdiffeq, not this repository here.

But if you want to work with this repository, proceed as follows.

Installation

We use Python 3.12 for all experiments. Older versions might also work. Ensure that JAX is installed. Then, run

pip install .

This command installs the source code plus all dependencies.

Working with the source

After following the installation instructions above, the test dependencies are installed.

To run the tests, run

make test

To format the code, run

make format-and-lint