Pith. sign in

REVIEW 4 major objections 5 minor 65 references

TimeGraph: Synthetic Benchmark Datasets for Robust Time-Series Causal Discovery

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read TimeGraph supplies synthetic time-series benchmark datasets with known ground-truth causal graphs, and its evaluations show that standard causal discovery algorithms collapse on nonlinear and confounded variants.

desk verdict Useful benchmark resource with a genuine gap to fill, but the SHD numbers in the evaluation are internally inconsistent and the abstract oversells the scope. read the letter →

arxiv 2506.01361 v1 pith:PPSNA4YS submitted 2025-06-02 cs.LG cs.IRstat.ML

Muhammad Hasan Ferdous , Emam Hossain , Md Osman Gani This is my paper
classification cs.LGcs.IRstat.ML
keywords causaldiscoverybenchmarkssynthetictimeseriestemporalcausalitygroundtruthdatasetsmultivariatenon-stationarydatamissinglatentconfounders
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper introduces TimeGraph, a family of synthetic time-series datasets built so that the true causal graph is always known. The datasets deliberately include realistic complications that most benchmarks leave out: trends, seasonality, irregular sampling, heavy-tailed or mixed noise, random and block missing data, and unobserved confounders. The central claim is that these complications change how well causal discovery algorithms perform, with standard methods that do well on linear-Gaussian benchmarks dropping to near-random performance on nonlinear or confounded variants. If the claim holds, TimeGraph provides a controlled testbed for locating exactly where algorithms fail and for comparing new methods on equal terms.

What carries the argument

The central object is the dataset variant generator, a time-lagged structural equation model with a known ground-truth causal graph. The generator combines four levers: functional form (linear, polynomial, or trigonometric trend-seasonal), error distribution (Gaussian, Student-t, or Gaussian-Laplace mixture), temporal sampling (regular or exponential waiting-time irregular), and missingness mechanism (none, MCAR, block, or combined), plus an optional latent confounder inserted into the equations. The key design move is paired construction: for each functional family the confounded and unconfounded versions share the same observable equations, so any performance gap between them can be attributed to the hidden variable. PCMCI+ and LPCMCI are not part of the generation machinery; they are evaluation probes that use conditional-independence and momentary-conditional-independence tests to try to recover the graph the generator produced.

What would settle it

Check the results table against the released ground-truth graphs: a 4-variable graph with lags 0, 1, and 2 has at most 48 directed edges, so any reported edge-error score above that bound, such as 45 or 51, would indicate the metric or graph representation is inconsistent with the stated protocol. If those scores do not reproduce from the released scripts, the claim that standard algorithms fail near-randomly on the nonlinear variants would need to be re-evaluated.

Watch

Extended reading notes

Core claim

TimeGraph is a benchmark suite of synthetic multivariate time series generated from known structural equation models with lagged dependencies. Each dataset family comes in two versions, one with all variables observed and one with an added latent confounder, so the effect of hidden confounding can be isolated while the ground-truth graph stays fixed. The functional forms progress from linear autoregressions to polynomial and trigonometric trend-seasonal models, with variants for irregular timestamps, mixed error distributions, MCAR and block missingness, and sample sizes from 500 to 5,000. The paper reports evaluations of FGES, PC, PCMCI+, and LPCMCI using true positive rate, false discovery rate, and structural Hamming distance, and finds that performance depends sharply on these temporal conditions: linear Gaussian datasets are recovered well, while purely nonlinear datasets drive true positive rate toward zero and false discovery rate toward one. The suite is released with generation scripts and evaluation protocols so that results can be reproduced.

Load-bearing premise

The benchmark's conclusions depend on the assumption that the scores used to compare algorithms are computed correctly against the true graphs, including the edge-error counts in the results tables.

Editorial extensions

If this is right

  • On the released linear-Gaussian variants, all four evaluated algorithms recover near-correct graphs, so these variants can serve as sanity checks for new methods.
  • On polynomial and trigonometric trend-seasonal variants, the same algorithms drop to near-zero true positive rate and near-one false discovery rate, so results on linear or stationary data cannot be extrapolated to realistic temporal settings.
  • Comparing each dataset family with its confounded counterpart isolates the effect of latent confounding, letting users test whether a method's robustness claim holds.
  • The irregular-sampling and missing-data variants provide controlled conditions for studying how preprocessing and imputation choices interact with causal discovery.
  • Because every dataset ships with a fully specified causal graph, true positive rate, false discovery rate, and structural Hamming distance can be computed uniformly across studies, which is what enables cross-paper comparison.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Extension: the paired confounded and unconfounded designs could support a single 'confounding robustness' score, such as the performance drop from the unconfounded to the confounded version of each family, which the paper reports only implicitly.
  • Extension: if the edge-error scores in the results table are inconsistent with the released ground-truth graphs, the paper's specific algorithm rankings should be treated as provisional; the benchmark artifacts themselves, not the rankings, would be the durable contribution.
  • Extension: the complete failure of all evaluated methods on the polynomial-Gaussian family suggests that variant may be too difficult to discriminate between algorithms; a graded nonlinearity ladder would be a natural next step for benchmark design.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper introduces TimeGraph, a synthetic benchmark suite for time-series causal discovery. It describes ten dataset families (A1, A1C, B1, B1C, B2, B2C, C1, C1C, C2, C2C, D1, D1C, D2, D2C, D3, D3C) that combine linear or nonlinear functional forms, Gaussian/Student-t/Laplace noise, regular or irregular sampling, MCAR/block/combined missingness, and optional latent confounders, each claimed to have a known ground-truth temporal causal graph. The authors release generation scripts and demonstrate the suite by evaluating FGES, PC, PCMCI+, and LPCMCI on a subset of the variants, reporting TPR, FDR, and SHD. The main claims are that TimeGraph provides a more realistic and systematic benchmark than existing synthetic resources and that the experiments reveal large performance variations under nonlinearity and confounding.

Significance. If the dataset generation and evaluation metrics are made precise and reproducible, TimeGraph could be a useful community resource: it combines trends, seasonality, irregular sampling, missingness, and latent confounding in a single controlled framework, and the public code release is a concrete strength. The proof-of-concept evaluation also highlights practically relevant failure modes of standard algorithms on nonlinear and confounded time series. However, the manuscript currently does not specify the generative equations or the exact metric protocol, and the reported SHD values are inconsistent with the stated graph size; the empirical illustration therefore needs to be repaired before the benchmark's utility claim is fully supported.

major comments (4)
  1. [Table 2 / Section 5.2] The reported SHD values are inconsistent with the stated graph size. For a dataset with 4 observed variables and tau_max=2, the standard Tigramite graph representation (used in Section 3, 'Implementation and Evaluation') has at most 4*4*3=48 directed link slots (44 if contemporaneous self-loops are excluded). Any SHD computed as the number of edge additions, deletions, and direction reversals over that representation is bounded by the number of slots, so SHD values of 45 (B1 with Student's t error, PCMCI+), 51 (B1C with Student's t error, PCMCI+), and 36 (B1C with Student's t error, LPCMCI) cannot arise under the stated protocol. If an unrolled 3-slice DAG encoding with 12 nodes was used instead, the effective number of variables and the TPR/FDR denominators change, and the paper does not state this. Because Section 5.2's conclusions about algorithm performance on nonlinear and confounded variants rest on this table, the evaluation must be recomputed and accompanied by an explicit definition of the graph encoding and of SHD.
  2. [Section 3 / Table 1] The generative models are described only verbally. The paper does not provide structural equations, coefficient values, base graph edge sets, graph densities, or parameter ranges for any of the variants, nor does it specify how the latent confounder U enters the equations and whether its incident edges are part of the ground-truth graph. As a result, the central claim that each dataset is accompanied by a fully specified causal graph cannot be verified from the manuscript without inspecting the code. A benchmark paper should include, at minimum in an appendix, the exact equations, the baseline graph adjacency matrices, and the parameter settings for each variant.
  3. [Section 5 / Abstract and Conclusion] The empirical evaluation is narrower than the paper's claims. Table 2 contains only ten rows, all with 4 variables, lag 2, and 500 samples, and only the A1, B1, and C1 families (with and without confounders) are tested. None of the irregular-sampling variants (A2, B2, C2), missing-data variants (D1, D2, D3), or mixed-error variants is evaluated, despite the abstract and conclusion stating that the experiments reveal performance variation under 'sampling irregularity' and 'missing data patterns.' The authors should either add experiments for those variants or restrict the empirical claims to the variants actually tested.
  4. [Section 5, metric definitions] The definitions of TPR, FDR, and SHD are ambiguous for the confounded variants. For A1C, B1C, and C1C, it is not stated whether the latent variable U is included in the ground-truth graph or whether the ground truth is a projection onto the observed variables; this changes the denominators of TPR/FDR and the set of edges over which SHD is computed. The paper defines the metrics only in words and does not specify counting rules for contemporaneous edges or for direction reversals. This ambiguity is part of the reason the Table 2 numbers cannot be checked from the manuscript alone.
minor comments (5)
  1. [Figure 2 / Section 4] The comparison between the C1 synthetic series and real stock data lacks axis labels, units, and a source for the real data, and the claimed 'qualitative similarities' are not quantified.
  2. [Section 4 / Table 1] The Gaussian noise is written as 'N(0, 0.12)' without stating whether 0.12 is a variance or a standard deviation; the Student-t degrees of freedom should also be stated consistently in Table 1 and Section 4.
  3. [Section 7 / Table 2] Section 7 says the evaluation operates at 'moderate scale (e.g., up to 8 variables),' but the only reported experimental results are for 4 variables; Figure 3 shows graphs for 6 and 8 variables but no corresponding results. The text should clarify what was actually run.
  4. [Section 5.1] The description of the PC baseline does not state how it was adapted to time-series data (e.g., lagged-variable expansion) or which conditional-independence test was used; similarly, the FGES scoring criterion is not specified.
  5. [Table 2] The dash notation for missing results is not explained, and some rows have formatting issues (e.g., missing spaces between the dataset label and the first metric column).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TimeGraph's synthetic data generation and algorithm evaluations are self-contained, with the SHD anomaly being a correctness concern rather than a circularity.

full rationale

The paper's central contribution is a set of synthetic time-series datasets with known ground-truth causal graphs; the ground truth is defined by explicit generation equations and graph encodings, not fitted to any outcome or evaluation metric. The empirical claim that algorithm performance varies under nonlinearity, confounding, trends, and missingness is an observed result of running FGES, PC, PCMCI+, and LPCMCI on those datasets, so it does not reduce to the inputs by construction. No parameter is fitted to the TPR/FDR/SHD numbers that are later reported, and no derived quantitative claim is definitionally equivalent to a fitted value. The self-citations in the paper ([15], [16], [25], [27], [28]) appear in related-work and background contexts and are not load-bearing for the benchmark construction or the evaluation conclusions. Section 7 explicitly acknowledges limitations such as synthetic nature, moderate scale, observational-only data, and simplified confounders; these are caveats about scope and realism, not circular premises. The main substantive concern is the internal consistency of Table 2, where SHD values such as 51 and 45 appear for datasets stated to have 4 variables and a maximum lag of 2, which exceeds the 48 link slots of the standard Tigramite 4x4x3 lagged-graph representation unless a different encoding is intended. That is an evaluation-correctness or reporting issue and could undermine the numeric conclusions, but it is not circular reasoning. Therefore the appropriate circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 1 invented entities

The central claim is that TimeGraph provides a useful benchmark resource; this rests on the correctness of the data generation and the evaluation metrics. The paper does not present explicit equations for all variants, so the implementation is assumed to match the description. The evaluation conclusion depends on the standard metrics TPR, FDR, and SHD, which are computed by the provided code. No fitted parameters are used to derive a quantitative result; the dataset parameters are design choices.

assumptions (3)
  • domain assumption Additive noise structural equation models with fixed lagged dependencies correctly represent the ground-truth causal graphs.
    Section 3 states datasets are generated from structural equations with fixed lagged dependencies; the evaluation assumes these equations realize the intended causal graph.
  • domain assumption Standard causal discovery evaluation metrics (TPR, FDR, SHD) computed on the time-lagged graph are valid measures of performance.
    Section 5 defines these metrics as the basis for comparing learned graphs to ground truth.
  • ad hoc to paper The provided source code faithfully implements the described data generation process.
    The paper does not include explicit equations for all variants, so verification of the datasets requires trusting or running the linked GitHub repository.
invented entities (1)
  • Latent confounder U
    purpose: Added to confounded dataset variants to create hidden common causes that affect multiple observed variables, used to test robustness of causal discovery algorithms.
    U is a simulation construct defined within the dataset generation; it is not a claim about any real-world entity and has no falsifiable handle outside the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TimeGraph: Synthetic Benchmark Datasets for Robust Time-Series Causal Discovery." pith.science (2026). https://pith.science/paper/PPSNA4YS

@misc{pith2026250601361,
  author       = {Pith},
  title        = {Pith review of: TimeGraph: Synthetic Benchmark Datasets for Robust Time-Series Causal Discovery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PPSNA4YS}},
  note         = {Machine review of arXiv:2506.01361}
}
read the original abstract

Robust causal discovery in time series datasets depends on reliable benchmark datasets with known ground-truth causal relationships. However, such datasets remain scarce, and existing synthetic alternatives often overlook critical temporal properties inherent in real-world data, including nonstationarity driven by trends and seasonality, irregular sampling intervals, and the presence of unobserved confounders. To address these challenges, we introduce TimeGraph, a comprehensive suite of synthetic time-series benchmark datasets that systematically incorporates both linear and nonlinear dependencies while modeling key temporal characteristics such as trends, seasonal effects, and heterogeneous noise patterns. Each dataset is accompanied by a fully specified causal graph featuring varying densities and diverse noise distributions and is provided in two versions: one including unobserved confounders and one without, thereby offering extensive coverage of real-world complexity while preserving methodological neutrality. We further demonstrate the utility of TimeGraph through systematic evaluations of state-of-the-art causal discovery algorithms including PCMCI+, LPCMCI, and FGES across a diverse array of configurations and metrics. Our experiments reveal significant variations in algorithmic performance under realistic temporal conditions, underscoring the need for robust synthetic benchmarks in the fair and transparent assessment of causal discovery methods. The complete TimeGraph suite, including dataset generation scripts, evaluation metrics, and recommended experimental protocols, is freely available to facilitate reproducible research and foster community-driven advancements in time-series causal discovery.

Figures

Figures reproduced from arXiv: 2506.01361 by the authors.

Figure 1
Figure 1. Illustrative time-series causal graph. Nodes denote [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Comparison of our synthetic time series (C1 variant, [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Combined linear causal graphs for nine configura [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Comparative bar charts illustrating the performance of four causal discovery algorithms (FGES, PC, PCMCI+, LPCMCI) [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

65 extracted references · 59 canonical work pages

  1. [1]

    Ossama Ahmed, Frederik Träuble, Anirudh Goyal, Alexander Neitz, Yoshua Ben- gio, Bernhard Schölkopf, Manuel Wüthrich, and Stefan Bauer. 2020. Causalworld: A robotic manipulation benchmark for causal structure and transfer learning. arXiv preprint arXiv:2010.04296 (2020)

    arXiv 2020

  2. [2]

    Torben G Andersen, Tim Bollerslev, Francis X Diebold, and Paul Labys. 2001. The distribution of realized exchange rate volatility. Journal of the American statistical association 96, 453 (2001), 42–55

    work page 2001

  3. [3]

    Angrist and Jörn-Steffen Pischke

    Joshua D. Angrist and Jörn-Steffen Pischke. 2009. Mostly Harmless Econometrics: An Empiricist’s Companion. Princeton University Press, Princeton, NJ

    work page 2009

  4. [4]

    Elias Bareinboim and Judea Pearl. 2016. Causal inference and the data-fusion problem. Proceedings of the National Academy of Sciences 113, 27 (2016), 7345– 7352

    work page 2016

  5. [5]

    Ben Bernanke. 2005. The global saving glut and the US current account deficit . Technical Report. Board of Governors of the Federal Reserve System (US)

    work page 2005

  6. [6]

    Olivier J Blanchard and Danny Quah. 1988. The dynamic effects of aggregate demand and supply disturbances

    work page 1988

  7. [7]

    Defu Cao, James Enouen, Yujing Wang, Xiangchen Song, Chuizheng Meng, Hao Niu, and Yan Liu. 2023. Estimating Treatment Effects from Irregular Time Series Observations with Hidden Confounders. Proceedings of the AAAI Conference on Artificial Intelligence 37, 6 (Jun. 2023), 6897–6905. doi:10.1609/aaai.v37i6.25844

    work page doi:10.1609/aaai.v37i6.25844 2023

  8. [8]

    Intergovernmental Panel On Climate Change. 2007. Climate change 2007: the physical science basis. Agenda 6, 07 (2007), 333

    work page 2007

Show all 65 references
  1. [9]

    Chris Chatfield and Haipeng Xing. 2019. The Analysis of Time Series: An Intro- duction with R. Chapman and Hall/CRC, Boca Raton, FL

    2019

  2. [10]

    Yonghong Chen, Govindan Rangarajan, Jianfeng Feng, and Mingzhou Ding

  3. [11]

    Yuxiao Cheng, Ziqian Wang, Tingxiong Xiao, Qin Zhong, Jinli Suo, and Kunlun He. 2023. Causaltime: Realistically Generated Time-Series for Benchmarking of Causal Discovery. arXiv preprint arXiv:2310.01753 (2023). arXiv:2310.01753

    2023 arXiv

  4. [12]

    David Maxwell Chickering. 2002. Optimal structure identification with greedy search. Journal of machine learning research 3, Nov (2002), 507–554

    2002

  5. [13]

    Rhian M Daniel, Bianca L De Stavola, Simon N Cousens, and Stijn Vansteelandt

  6. [14]

    Eugene F Fama. 1970. Efficient capital markets. Journal of finance 25, 2 (1970), 383–417

    1970

  7. [15]

    Muhammad Hasan Ferdous, Uzma Hasan, and Md Osman Gani. 2023. Cdans: Temporal Causal Discovery from Autocorrelated and Non-Stationary Time Se- ries Data. In Proceedings of the Machine Learning for Healthcare Conference , Vol. 216. PMLR, Online, 186–207. https://proceedings.mlr....

    2023

  8. [16]

    Muhammad Hasan Ferdous, Uzma Hasan, and Md Osman Gani. 2023. Ecdans: Ef- ficient Temporal Causal Discovery from Autocorrelated and Non-Stationary Data (Student Abstract). Proceedings of the AAAI Conference on Artificial Intelligence 37, 13 (2023), 16208–16209

    2023

  9. [17]

    Alexander Gain and Ilya Shpitser. 2018. Structure Learning under Missing Data. In Proceedings of the International Conference on Probabilistic Graphical Models (PGM), Vol. 72. PMLR, Aalborg, Denmark, 121–132. https://proceedings.mlr. press/v72/gain18a.html

    2018

  10. [18]

    João Gama, Indr˙e Žliobait˙e, Albert Bifet, Mykola Pechenizkiy, and Abdelhamid Bouchachia. 2014. A survey on concept drift adaptation. ACM computing surveys (CSUR) 46, 4 (2014), 1–37

    2014

  11. [19]

    Md Osman Gani, Shravan Kethireddy, Riddhiman Adib, Uzma Hasan, Paul Grif- fin, and Mohammad Adibuzzaman. 2023. Structural causal model with expert augmented knowledge to estimate the effect of oxygen therapy on mortality in the icu. Artificial intelligence in medicine 137 (202...

    2023

  12. [20]

    Andreas Gerhardus and Jakob Runge. 2020. High-recall causal discovery for au- tocorrelated time series with latent confounders. Advances in neural information processing systems 33 (2020), 12615–12625

    2020

  13. [21]

    Marzyeh Ghassemi, Marco Pimentel, Tristan Naumann, Thomas Brennan, David Clifton, Peter Szolovits, and Mengling Feng. 2015. A Multivariate Time Series Modeling Approach to Severity of Illness Assessment and Forecasting in ICU with Sparse, Heterogeneous Clinical Data. Proceedin...

    2015

  14. [22]

    Chang Gong, Chuzhe Zhang, Di Yao, Jingping Bi, Wenbin Li, and YongJun Xu

  15. [23]

    Clive W. J. Granger. 1969. Investigating Causal Relations by Econometric Models and Cross-Spectral Methods. Econometrica 37, 3 (1969), 424–438

    1969

  16. [24]

    Hamilton

    James D. Hamilton. 2020. Time Series Analysis . Princeton University Press, Princeton, NJ

    2020

  17. [25]

    Uzma Hasan, Emam Hossain, and Md Osman Gani. 2023. A Survey on Causal Discovery Methods for I.I.D. and Time Series Data. Transactions on Machine Learning Research (2023). https://openreview.net/forum?id=YdMrdhGx9y

    2023

  18. [26]

    Miguel A Hernán and James M Robins. 2010. Causal inference

    2010

  19. [27]

    Subramanian, and Md Osman Gani

    Emam Hossain, Sahara Ali, Yiyi Huang, Nicole-Jeanne Schlegel, Jianwu Wang, Aneesh C. Subramanian, and Md Osman Gani. 2024. Incorporating Causality with Deep Learning in Predicting Short-Term and Seasonal Sea Ice. Presented at the 104th AMS Annual Meeting. Baltimore, MD, Januar...

    2024

  20. [28]

    Emam Hossain, Muhammad Hasan Ferdous, Jianwu Wang, Aneesh Subramanian, and Md Osman Gani. 2025. Correlation to Causation: A Causal Deep Learning Framework for Arctic Sea Ice Prediction. arXiv preprint arXiv:2503.02093 (2025)

    2025 arXiv

  21. [29]

    Biwei Huang, Kun Zhang, Jiji Zhang, Joseph Ramsey, Ruben Sanchez-Romero, Clark Glymour, and Bernhard Schölkopf. 2020. Causal discovery from heteroge- neous/nonstationary data. Journal of Machine Learning Research 21, 89 (2020), 1–53

    2020

  22. [30]

    Samuel Kotz, Tomasz Kozubowski, and Krzystof Podgorski. 2012. The Laplace Distribution and Generalizations: A Revisit with Applications to Communications, Economics, Engineering, and Finance . Springer Science & Business Media, New York, NY

    2012

  23. [31]

    Stuart R Lipsitz, Kyungmann Kim, and Lueping Zhao. 1994. Analysis of repeated categorical data using generalized estimating equations. Statistics in medicine 13, 11 (1994), 1149–1163

    1994

  24. [32]

    Roderick J. A. Little and Donald B. Rubin. 2019. Statistical Analysis with Missing Data. John Wiley & Sons, Hoboken, NJ

    2019

  25. [33]

    Daniel Malinsky and Peter Spirtes. 2019. Learning the Structure of a Nonstation- ary Vector Autoregression. In Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS), Vol. 89. PMLR, Naha, Okinawa, Japan, 2986–2994. https://proceedi...

    2019

  26. [34]

    Michael E Mann, Raymond S Bradley, and Malcolm K Hughes. 1999. Northern hemisphere temperatures during the past millennium: Inferences, uncertainties, and limitations. Geophysical research letters 26, 6 (1999), 759–762

    1999

  27. [35]

    Joris M Mooij, Jonas Peters, Dominik Janzing, Jakob Zscheischler, and Bernhard Schölkopf. 2016. Distinguishing cause from effect using observational data: methods and benchmarks. Journal of Machine Learning Research 17, 32 (2016), 1–102

    2016

  28. [36]

    National Research Council, Committee on National Statistics, and Panel on Handling Missing Data in Clinical Trials. 2011. The Prevention and Treatment of Missing Data in Clinical Trials . Technical Report. National Academies Press, Washington, DC

    2011

  29. [37]

    Judea Pearl. 2009. Causality. Cambridge University Press, Cambridge, UK

    2009

  30. [38]

    Jonas Peters, Dominik Janzing, and Bernhard Schölkopf. 2013. Causal Inference on Time Series Using Restricted Structural Equation Models. In Advances in Neural Information Processing Systems (NeurIPS) , Vol. 26. Curran Associates, Inc., Red Hook, NY, 154–162

    2013

  31. [39]

    Jonas Peters, Dominik Janzing, and Bernhard Schölkopf. 2017. Elements of Causal Inference: Foundations and Learning Algorithms. The MIT Press, Cambridge, MA

    2017

  32. [40]

    Adrian E Raftery, Steven Lewis, et al. 1992. How many iterations in the Gibbs sampler. Bayesian statistics 4, 2 (1992), 763–773

    1992

  33. [41]

    Joseph Ramsey, Madelyn Glymour, Ruben Sanchez-Romero, and Clark Glymour

  34. [42]

    CLEVELAND RB. 1990. STL: A seasonal-trend decomposition procedure based on loess. J Off Stat 6 (1990), 3–73

    1990

  35. [43]

    Roberto Refinetti. 2016. Circadian Physiology. CRC Press, Boca Raton, FL

    2016

  36. [44]

    Alex Reinhart. 2018. A review of self-exciting spatio-temporal point processes and their applications. Statist. Sci. 33, 3 (2018), 299–318

    2018

  37. [45]

    James M Robins, Andrea Rotnitzky, and Lue Ping Zhao. 1995. Analysis of semi- parametric regression models for repeated outcomes in the presence of missing data. Journal of the american statistical association 90, 429 (1995), 106–121

    1995

  38. [46]

    Donald B Rubin. 1976. Inference and missing data. Biometrika 63, 3 (1976), 581–592

    1976

  39. [47]

    Jakob Runge. 2020. Discovering Contemporaneous and Lagged Causal Relations in Autocorrelated Nonlinear Time Series Datasets. In Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI) , Vol. 124. PMLR, Online, 1388–1397. https://proceedings.mlr.press...

    2020

  40. [48]

    Jakob Runge, Peer Nowack, Marlene Kretschmer, Seth Flaxman, and Dino Sejdi- novic. 2019. Detecting and quantifying causal associations in large nonlinear time series datasets. Science advances 5, 11 (2019), eaau4996

    2019

  41. [49]

    Karen Sachs, Omar Perez, Dana Pe’er, Douglas A Lauffenburger, and Garry P Nolan. 2005. Causal protein-signaling networks derived from multiparameter single-cell data. Science 308, 5721 (2005), 523–529

    2005

  42. [50]

    Peter Schulam and Suchi Saria. 2017. Reliable Decision Support Using Counter- factual Models. In Advances in Neural Information Processing Systems (NeurIPS) , Vol. 30. Curran Associates, Inc., Red Hook, NY, 1697–1708. https://dl.acm.org/ doi/pdf/10.5555/3294771.3294933

    2017

  43. [51]

    Hoyer, Aapo Hyvärinen, Antti Kerminen, and Michael I

    Shohei Shimizu, Patrik O. Hoyer, Aapo Hyvärinen, Antti Kerminen, and Michael I. Jordan. 2006. A Linear Non-Gaussian Acyclic Model for Causal Discovery.Journal of Machine Learning Research 7 (2006), 2003–2030

    2006

  44. [52]

    Shumway and David S

    Robert H. Shumway and David S. Stoffer. 2017. Time Series Analysis and Its Applications: With R Examples (4 ed.). Springer, New York, NY. TimeGraph: Synthetic Benchmark Datasets for Robust Time-Series Causal Discovery KDD ’25, August 3–7, 2025, Toronto, ON, Canada

    2017

  45. [53]

    Peter Spirtes, Clark Glymour, and Richard Scheines. 2000. Causation, Prediction, and Search (2 ed.). MIT Press, Cambridge, MA

    2000

  46. [54]

    James H Stock and Mark W Watson. 2002. Macroeconomic forecasting using diffusion indexes. Journal of Business & Economic Statistics 20, 2 (2002), 147–162

    2002

  47. [55]

    Ruibo Tu, Cheng Zhang, Paul Ackermann, Karthika Mohan, Hedvig Kjellström, and Kun Zhang. 2019. Causal Discovery in the Presence of Missing Data. In Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics (AISTATS) (Proceedings of Mac...

    2019

  48. [56]

    Zeyu Wang. 2024. CausalBench: A Comprehensive Benchmark for Evaluating Causal Reasoning Capabilities of Large Language Models. In Proceedings of the 10th SIGHAN Workshop on Chinese Language Processing (SIGHAN-10). Association for Computational Linguistics, Bangkok, Thailand, 1...

    2024

  49. [57]

    Sascha Xu, Osman A Mian, Alexander Marx, and Jilles Vreeken. 2022. Inferring Cause and Effect in the Presence of Heteroscedastic Noise. In Proceedings of the 39th International Conference on Machine Learning (ICML) , Vol. 162. PMLR, Baltimore, MD, USA, 24615–24630. https://pro...

    2022

  50. [58]

    Hsiang-Fu Yu, Nikhil Rao, and Inderjit S. Dhillon. 2016. Temporal Regularized Matrix Factorization for High-Dimensional Time Series Prediction. In Advances in Neural Information Processing Systems (NeurIPS) , Vol. 29. Curran Associates, Inc., Red Hook, NY, 847–855. https://dl....

    2016

  51. [59]

    Ramsey, Kayhan Batmanghelich, Peter Spirtes, and Clark Glymour

    Kun Zhang, Mingming Gong, Joseph D. Ramsey, Kayhan Batmanghelich, Peter Spirtes, and Clark Glymour. 2018. Causal Discovery with Linear Non-Gaussian Models under Measurement Error: Structural Identifiability Results. InProceedings of the 34th Conference on Uncertainty in Artifi...

    2018

  52. [60]

    Kun Zhang and Aapo Hyvärinen. 2009. On the Identifiability of the Post- Nonlinear Causal Model. In Proceedings of the Twenty-Fifth Conference on Uncer- tainty in Artificial Intelligence (UAI). AUAI Press, Montreal, QC, Canada, 647–655. https://www.auai.org/uai2009/papers/UAI20...

    2009

  53. [61]

    Kun Zhang, Jonas Peters, Dominik Janzing, and Bernhard Schölkopf. 2011. Kernel- based Conditional Independence Test and Application in Causal Discovery. In Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence (UAI) . AUAI Press, Barcelona, Spain, 804–813

    2011

  54. [2004]

    Physics letters A 324, 1 (2004), 26–35

    Analyzing multiple nonlinear time series with extended Granger causality. Physics letters A 324, 1 (2004), 26–35

    2004

  55. [2015]

    Biometrics 71, 1 (2015), 1–14

    Causal mediation analysis with multiple mediators. Biometrics 71, 1 (2015), 1–14

    2015

  56. [2017]

    International journal of data science and analytics 3 (2017), 121–129

    A million variables and more: the fast greedy equivalence search algorithm for learning high-dimensional graphical causal models, with an application to functional magnetic resonance images. International journal of data science and analytics 3 (2017), 121–129

    2017

  57. [2024]

    Causal discovery from temporal data: An overview and new perspectives. Comput. Surveys 57, 4 (2024), 1–38

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.