kausable/CausalDynamics

1---2license: mit3---4 5 6# CausalDynamics: A large-scale benchmark for structural discovery of dynamical causal models7 8<div align="center">9  <h3>NeurIPS 2025</h3>10</div>11 12 13<div align="center">14<a href="https://arxiv.org/abs/2505.16620"><img src="https://img.shields.io/badge/arXiv-2505.16620-b31b1b.svg" alt="arXiv"/></a>15<a href="http://kausable.github.io/CausalDynamics"><img src="https://img.shields.io/badge/View-Documentation-blue?style=for-the-badge)" alt="Homepage"/></a>16  <!-- <a href="<ADD_LINK>"><img src="https://img.shields.io/badge/ArXiV-2402.00712-b31b1b.svg" alt="arXiv"/></a> -->17<a href="https://huggingface.co/datasets/kausable/CausalDynamics"><img src="https://img.shields.io/badge/Dataset-HuggingFace-ffd21e" alt="Huggingface Dataset"/></a>18<a href="https://github.com/kausable/CausalDynamics/blob/main/LICENSE.txt"><img src="https://img.shields.io/badge/License-MIT-green" alt="License Badge"/></a>19<a href="https://github.com/kausable/CausalDynamics/actions/workflows/run-tests.yml"><img src="https://github.com/kausable/CausalDynamics/workflows/Run%20Tests/badge.svg" alt="Tests"/></a>20</div>21</br>22 23A comprehensive benchmark framework designed to rigorously evaluate state-of-the-art causal discovery algorithms for dynamical systems.24 25## Key Features26 271️⃣ **Large-Scale Benchmark**. Systematically evaluate state-of-the-art causal discovery algorithms on thousands of graph challenges with increasing difficulty. 28 292️⃣ **Customizable Data Generation**. Scalable, user-friendly generation of increasingly complex coupled ordinary and stochastic systems of differential equations30 313️⃣ **Diverse Challenges**. From simple chaotic systems to modular causal coupling of dynamical systems, including optional noise, confounding, time lags, and even climate model dynamics.32 33**Abstract**: Causal discovery for dynamical systems poses a major challenge in fields where active interventions are infeasible. Most methods used to investigate these systems and their associated benchmarks are tailored to deterministic, low-dimensional and weakly nonlinear time-series data. To address these limitations, we present *CausalDynamics*, a large-scale benchmark and extensible data generation framework to advance the structural discovery of dynamical causal models. Our benchmark consists of true causal graphs derived from thousands of coupled ordinary and stochastic differential equations as well as two idealized climate models. We perform a comprehensive evaluation of state-of-the-art causal discovery algorithms for graph reconstruction on systems with noisy, confounded, and lagged dynamics. *CausalDynamics* consists of a plug-and-play, build-your-own coupling workflow that enables the construction of a hierarchy of physical systems. We anticipate that our framework will facilitate the development of robust causal discovery algorithms that are broadly applicable across domains while addressing their unique challenges. 34 35## Datasets36You can generate your own dataset (see [getting started](#getting-started)), but you can also download our preprocessed ones directly from HuggingFace:37```bash38wget https://huggingface.co/datasets/kausable/CausalDynamics/resolve/main/process_causaldynamics.py39python process_causaldynamics.py40```41 42## Installation43### Using pip44 45*CausalDynamics* is available on [PyPi](https://pypi.org/project/causaldynamics/), so you can use pip to install `causaldynamics`, which currently requires Python version `3.10`.46 47```bash48pip install causaldynamics49```50 51If you use conda, please use the following commands: 52```shell53conda create --name venv python=3.1054conda activate venv55pip install causaldynamics56```57 58### Using pdm59Clone the repository and install it using [pdm](https://pdm-project.org/en/latest/): 60 61```shell62git clone https://github.com/kausable/CausalDynamics.git63cd CausalDynamics64pdm install65```66 67You can test whether the installation succeded by creating some coupled causal model data:68 69```shell70$(pdm venv activate)71python src/causaldynamics/creator.py --config config.yaml72```73 74You find the output at `output/<timestamp>` as default location.75 76 77## Getting Started78 79- [Challenge](https://kausable.github.io/CausalDynamics/challenge.html)80- [Quickstart](https://kausable.github.io/CausalDynamics/notebooks/quickstart.html)81- [CausalDynamics](https://kausable.github.io/CausalDynamics/notebooks/causaldynamics.html)82- [Benchmark](https://kausable.github.io/CausalDynamics/benchmark.html)83    - [Simple Dynamics](https://kausable.github.io/CausalDynamics/notebooks/simple_causal_models.html)84    - [Coupled Dynamics](https://kausable.github.io/CausalDynamics/notebooks/coupled_causal_models.html)85    - [Climate Dynamics](https://kausable.github.io/CausalDynamics/notebooks/climate_causal_models.html)86- [Troubleshoot](https://kausable.github.io/CausalDynamics/troubleshoot.html)87 88## Benchmarking89- [Baseline](https://kausable.github.io/CausalDynamics/baseline.html)90- [Evaluation](https://kausable.github.io/CausalDynamics/notebooks/eval_pipeline.html)91- [Leaderboard](https://kausable.github.io/CausalDynamics/leaderboard.html)92 93## Citation94If you find any of the code and dataset useful, feel free to acknowledge our work through:95 96```bibtex97@article{herdeanu2025causaldynamics,98  title={CausalDynamics: A large-scale benchmark for structural discovery of dynamical causal models},99  author={Herdeanu, Benjamin and Nathaniel, Juan and Roesch, Carla and Buch, Jatan and Ramien, Gregor and Haux, Johannes and Gentine, Pierre},100  journal={arXiv preprint arXiv:2505.16620},101  year={2025}102}103```