mlsanityData to 14 Oct 2026

Better Simulations for Validating Causal Discovery with the DAG-Adaptation of the Onion Method

arXiv 2405.13100 ↗First posted 21 May 2024stat.ME cs.AIcited by 6

Not on the map. Why?

Abstract (arXiv)

The number of artificial intelligence algorithms for learning causal models from data is growing rapidly. Most ``causal discovery'' or ``causal structure learning'' algorithms are primarily validated through simulation studies. However, no widely accepted simulation standards exist and publications often report conflicting performance statistics -- even when only considering publications that simulate data from linear models. In response, several manuscripts have criticized a popular simulation design for validating algorithms in the linear case. We propose a new simulation design for generating linear models for directed acyclic graphs (DAGs): the DAG-adaptation of the Onion (DaO) method. DaO simulations are fundamentally different from existing simulations because they prioritize the distribution of correlation matrices rather than the distribution of linear effects. Specifically, the DaO method uniformly samples the space of all correlation matrices consistent with (i.e. Markov to) a DAG. We also discuss how to sample DAGs and present methods for generating DAGs with scale-free in-degree or out-degree. We compare the DaO method against two alternative simulation designs and provide implementations of the DaO method in Python and R: https://github.com/bja43/DaO_simulation. We advocate for others to adopt DaO simulations as a fair universal benchmark.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 22 of its 70 references that are in the map. First posted on arXiv 21 May 2024.

Builds on

22 of 70 references.

References · 22

  1. 2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 752
  2. 2021Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Gamecited by 161
  3. 2022DAGMA: Learning DAGs via M-matrices and a Log-Determinant Acyclicity Characterizationcited by 140
  4. 2013Causal Networks: Semantics and Expressivenesscited by 67
  5. 2022Greedy Relaxations of the Sparsest Permutation Algorithmcited by 64
  6. 2021Unsuitability of NOTEARS for Causal Graph Discoverycited by 47
  7. 2013Strong Completeness and Faithfulness in Bayesian Networkscited by 45
  8. 2013Parameter Priors for Directed Acyclic Graphical Models and the Characterization of Several Probability Distributionscited by 42
  9. 2023Fast Scalable and Accurate Discovery of DAGs Using the Best Order Score Search and Grow-Shrink Treescited by 38
  10. 2023Structure Learning with Continuous Optimization: A Sober Look and Beyondcited by 36
  11. 2014Addendum on the scoring of Gaussian directed acyclic graphical modelscited by 33
  12. 2019Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluationcited by 32
10 more
  1. 2018Efficient Sampling and Structure Learning of Bayesian Networkscited by 28
  2. 2023A Scale-Invariant Sorting Criterion to Find a Causal Order in Additive Noise Modelscited by 28
  3. 2023Assumption violations in causal discovery and the robustness of score matchingcited by 28
  4. 2021Data Generating Process to Evaluate Causal Discovery Techniques for Time Series Datacited by 22
  5. 2021Bayesian structure learning and sampling of Bayesian networks with the R package BiDAGcited by 17
  6. 2021Benchpress: A Scalable and Versatile Workflow for Benchmarking Structure Learning Algorithmscited by 11
  7. 2015A Uniformly Consistent Estimator of Causal Effects under the k-Triangle-Faithfulness Assumptioncited by 10
  8. 2016algcomparison: Comparing the Performance of Graphical Structure Learning Algorithms with TETRADcited by 3
  9. 2017A Comparison of Public Causal Search Packages on Linear, Gaussian Data with No Latent Variablescited by 2
  10. 2021A Uniformly Consistent Estimator of non-Gaussian Causal Effects Under the k-Triangle-Faithfulness Assumptioncited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 2; 2025: 3; 2026: 1.

6 papers.

  1. 2024The Landscape of Causal Discovery Data: Grounding Causal Discovery in Real-World Applicationscited by 18
  2. 2025Unitless Unrestricted Markov-Consistent SCM Generation: Better Benchmark Datasets for Causal Discoverycited by 8
  3. 2025Embracing Discrete Search: A Reasonable Approach to Causal Structure Learningcited by 4
  4. 2024Choosing DAG Models Using Markov and Minimal Edge Count in the Absence of Ground Truthcited by 1
  5. 2025An extensive simulation study evaluating the interaction of resampling techniques across multiple causal discovery contextscited by 0
  6. 2026A Topological Sorting Criterion for Random Causal Directed Acyclic Graphscited by 0