AutoCause: A Python framework that automates expert decisions in environmental time-series causal discovery

2026-07-31

AI Summary

This study addresses the challenges of inconsistent expert decisions in causal discovery from environmental time series, which often lead to incomparable, irreproducible, and unauditable results. To overcome this, we propose the first auditable Python workflow that explicitly formalizes and automates the entire causal discovery pipeline. The framework integrates four major causal discovery algorithms, non-causal baselines, conditional independence tests with multiple testing correction, and introduces a causal auditing module alongside a consensus-based edge scoring mechanism derived from multiple methods. Experiments across 145 datasets demonstrate that consensus-supported causal links achieve higher precision on synthetic benchmarks, and the overall framework substantially enhances the reliability, transparency, and reproducibility of causal analysis in environmental time series.

Abstract

Environmental time-series causal discovery requires expert decisions about method choice, conditional-independence tests, lag horizons, sample-size adequacy, multiple-testing control, and evidence interpretation. Applied inconsistently across datasets, these choices yield graphs that cannot be compared, reproduced, or audited. We present AutoCause, an open-source Python workflow that records each decision, derives defaults from an extended causal-audit module, and admits domain-informed overrides. The workflow wraps four established causal-discovery methods from three families, adds non-causal reference models, and grades links by method-count support. On 145 datasets from DGP-Atlas, TimeGraph, and a topology-derived CausalRivers reference, the methods recover complementary parts of the reference graphs. Majority-supported links are more precise than single-method links on the synthetic benchmarks but not against river topology. AutoCause converts inconsistent expert practice into an auditable, repeatable analysis; causal interpretation remains with the analyst. Available at https://github.com/marcoruizrueda/autocause.

Problem

• causal discovery
• environmental time-series
• expert decisions
• reproducibility
• auditability

Innovation

• causal discovery
• time-series analysis
• automated workflow
• method consensus
• reproducibility