Causal-TS: A Python Library for Causal Discovery in High-Dimensional and Nonstationary Time Series

Mohammad Fesanghary
7/27/2026
cs.LGstat.ME

Official code: bloomberg/causal-ts (1 star, GPL-3.0, last commit 2026-07-28).

Abstract

We describe Causal-TS, an open-source Python library for causal discovery in high-dimensional and nonstationary multivariate time series. Causal-TS provides four specialized algorithms-CDNOTS, CDNOTS+, CEDAR, and GRACE-along with wrappers for GES, Granger, LASSO-VAR, and LGES, all sharing a unified conditional independence (CI) test layer with GPU acceleration via PyTorch. A regime discovery pipeline detects structural breaks via pluggable changepoint detectors and runs discovery per regime with regime-specific parameters. A command-line interface, synthetic data generators, and optional DoWhy integration provide an end-to-end pipeline from raw time series to causal effect estimates. The library is pip-installable, tested on Python 3.10--3.12, and available at https://github.com/bloomberg/causal-ts.

View on arXivView PDF

Code Implementations(1)

bloomberg/causal-tsOfficial100%
GPL-3.0

Causal Discovery for Time Series

10Python, Jupyter NotebookJun 29, 20262 months agoGPL-3.0
research

Similar Papers

TTCD:Transformer Integrated Temporal Causal Discovery from Non-Stationary Time Series Data

85% match2026

Causal Discovery on Irregular Time Series

83% match2026

CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discovery

83% match1 repo1612023

Causal Discovery with Inverted Self-attention for Multivariate Time Series

82% match2026

Salesforce CausalAI Library: A Fast and Scalable Framework for Causal Analysis of Time Series and Tabular Data

82% match1 repo3242023

A Survey on Causal Discovery Methods for I.I.D. and Time Series Data

82% match2023

Cite this paper

@article{fesanghary2026causalts,
  title  = {Causal-TS: A Python Library for Causal Discovery in High-Dimensional and Nonstationary Time Series},
  author = {Mohammad Fesanghary},
  year   = {2026},
  eprint = {2607.24673},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2607.24673v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2607.24673.svg)](https://codeofpaper.com/paper/2607.24673)

Discussion