CausalRivers - Scaling up benchmarking of causal discovery for real-world time-series

Gideon Stein, Maha Shadaydeh, Jan Blunk, Niklas Penzel, Joachim Denzler

Keywords: Causal Discovery, Benchmarking, Time-series

TL;DR: The largest real-world causal discovery benchmark to this date, including high-resolution ts and ground truth causal graphs with over 1000 nodes.

Abstract:

Causal discovery, or identifying causal relationships from observational data, is a notoriously challenging task, with numerous methods proposed to tackle it.
Despite this, in-the-wild evaluation is still lacking, as works frequently rely on synthetic data evaluation and sparse real-world examples under critical theoretical assumptions.
Real-world causal structures, however, are often complex, evolving over time, non-linear, and influenced by unobserved factors, making
it hard for practitioners to select appropriate methods.
To bridge this gap, we introduce CausalRivers, the largest in-the-wild causal discovery benchmarking kit for time series data to date.
CausalRivers features an extensive dataset on river discharge that covers the complete eastern German territory (666 measurement stations) and the state of Bavaria (494 measurement stations).
It spans the years 2019 to 2023 with a 15-minute temporal resolution.
Further, we provide data from a recent flood around the Elbe River, as an event with a pronounced distributional shift.
Leveraging multiple sources of information and time-series meta-data, we constructed two distinct causal ground truth graphs (Bavaria and eastern Germany).
These graphs can be sampled to generate thousands of subgraphs to benchmark causal discovery across diverse and challenging settings.
To demonstrate the utility of our benchmarking kit, we evaluate several causal discovery approaches through multiple experiments and introduce effective baselines, identifying several areas for enhancement.
CausalRivers has the potential to facilitate robust evaluations and comparisons of causal discovery methods.
Besides this primary purpose, we also expect that this dataset will be relevant for connected areas of research, such as time series forecasting and anomaly detection.
Based on this, we hope to establish benchmark-driven method development that fosters advanced techniques for causal discovery, as is the case for many other areas of machine learning.

Primary Area: datasets and benchmarks

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 10051