Pith. sign in

REVIEW 3 major objections 5 minor 50 references

Generalized Steppingstone Sampling: Efficient marginal likelihood estimation in gravitational wave analysis of Pulsar Timing Array data

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Generalized steppingstone sampling estimates pulsar-timing marginal likelihoods cheaply and, applied to published data, strengthens the evidence for a gravitational-wave background.

desk verdict Useful application of a known estimator to PTA evidence, but the headline EPTA Bayes-factor increase rests on unverified reference-distribution coverage and a missing direct comparison to the original numbers. read the letter →

arxiv 2411.14736 v2 pith:XTVVV2YB submitted 2024-11-22 astro-ph.IM

El Mehdi Zahraoui , Patricio Maturana-Russel , Willem van Straten , Renate Meyer , Sergei Gulyaev This is my paper · ORCID
classification astro-ph.IM
keywords gravitationalwavebackgroundpulsartimingarraymarginallikelihoodestimationgeneralizedsteppingstonesamplingBayesfactorHellings-Downscorrelationcommonuncorrelatedrednoisemodelselection
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

Generalized steppingstone sampling (GSS) estimates marginal likelihoods in pulsar-timing-array Bayesian analyses at much lower computational cost than thermodynamic integration or classic steppingstone sampling, without losing accuracy. The paper's innovation is to start the sampling path from a reference distribution that approximates the posterior, so a handful of samples per rung and as few as eight temperature rungs suffice. Re-running published analyses, the method reproduces the 15-year North American evidence for a gravitational-wave background and sharpens its uncertainty, while the European second data release shows substantially stronger support for the Hellings-Downs correlation (the quadrupolar inter-pulsar correlation signature of a gravitational-wave background) over a common uncorrelated red-noise process, with log Bayes factors between 5.70 and 8.24 across the four data subsets. If the estimates are right, existing pulsar timing data already favor a gravitational-wave background more strongly than previously reported, and expensive spatial-correlation models can be compared directly and routinely.

What carries the argument

The central object is the generalized steppingstone estimator with its reference distribution $\pi_0$: instead of integrating from prior to posterior, GSS defines power posteriors $q_\beta = [L(X|\theta,M)\pi(\theta|M)]^\beta[\pi_0(\theta|M)]^{1-\beta}$ and writes the marginal likelihood as a product of importance-sampling ratios $r_k = E_{p_{\beta_{k-1}}}[(L\pi/\pi_0)^{\beta_k-\beta_{k-1}}]$. The reference distribution is built from a short posterior run as a product of independent Normal distributions fitted to each parameter's posterior mean and variance. Because $\pi_0$ already sits close to the posterior, the path between $\beta=0$ and $\beta=1$ is short, so only $K=8$--$16$ rungs with effective sample sizes of roughly 10--50 per rung are needed for stable estimates, and replicate runs give empirical Bayes-factor uncertainties. The temperature schedule uses Beta(0.3,1) quantiles to concentrate rungs near $\beta=0$, where adjacent power posteriors differ most.

What would settle it

Compute a nested-sampling or high-resolution thermodynamic-integration marginal likelihood for the same model and data subset that gave log BF approximately 8.24 (the European 'new' data set); if the independent estimate does not agree, or if the GSS importance weights for the spectral-index parameter show unbounded variance, then the reported increase in evidence is an artifact of the reference distribution.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that GSS provides accurate log-evidence estimates for pulsar-timing-array models with a small fraction of the usual sampling effort, and that this changes the reported evidence for the gravitational-wave background. In a 50-dimensional Gaussian benchmark with known $\log z = -115.38$, GSS converges with $K=4$ $\beta$ chains and ten samples per chain, while thermodynamic integration needs more than 32 chains and steppingstone sampling needs 16. For the 15-year North American data set, GSS gives $\log\mathrm{BF}_{\mathrm{HD/CURN}} = 5.11 \pm 1.24$, consistent with the original estimate $5.42 \pm 4.25$ but with tighter uncertainty. For the European second data release, GSS finds the gravitational-wave background favored over common uncorrelated red noise in every subset, with $\log\mathrm{BF}_{\mathrm{GWB/CURN}}$ from 5.70 to 8.24, which the paper reports as a substantial increase over the evidence originally published for that data set. The same machinery also yields single-pulsar noise-model selection through an inclusion Bayes factor computed from the GSS marginal-likelihood estimates.

Load-bearing premise

The method's cheap-and-accurate result depends on the approximate starting distribution having enough probability mass over the tails of every intermediate distribution along the sampling path, something the paper does not directly verify.

Editorial extensions

If this is right

  • Common-process models such as the Hellings-Downs correlation and overlap-reduction-function models can be compared directly by log Bayes factor, without the nested-model restrictions or arbitrary weighting of hypermodel sampling.
  • For the 15-year North American data, the gravitational-wave background is favored over a common uncorrelated red process with $\log\mathrm{BF} = 5.11\pm1.24$, consistent with the original analysis but with a standard deviation about four times smaller.
  • For all four European second-data-release subsets, the gravitational-wave background model is favored over common uncorrelated red noise with log Bayes factors from 5.70 to 8.24, while a binned overlap reduction function is not favored.
  • The inclusion Bayes factor, computed from GSS estimates, separates red-noise and dispersion-measure Gaussian-process contributions in single-pulsar analyses, giving a direct criterion for deciding whether to include noise parameters.
  • Because the reweighting method is a special case ($K=2$) of GSS and standard steppingstone sampling corresponds to choosing the prior as the reference distribution, the framework unifies several existing marginal-likelihood estimators.

Reading between the lines

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

  • If the GSS estimates are unbiased, the original European analysis underestimated the gravitational-wave evidence; a direct check would be to re-run those four data subsets with an independent nested-sampling or high-resolution thermodynamic-integration estimator and compare the log Bayes factors.
  • The main risk is reference-distribution under-coverage: heavy-tailed spectral-index or amplitude posteriors could make the ratios in Equation (14) high-variance even when the reported Monte Carlo standard errors are small, so reporting effective sample sizes and weight distributions for each $\beta$ rung would make this testable.
  • The cost structure suggests a workflow where a short posterior run calibrates $\pi_0$ and GSS then monitors evidence as new pulsars or new data releases arrive, turning marginal-likelihood estimation into a routine part of pulsar-timing-array operations.
  • The inclusion-Bayes-factor idea could be extended to other parameter blocks in common-signal models, such as solar-system-ephemeris or clock-noise terms, not just single-pulsar noise parameters.
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

3 major / 5 minor

Summary. The paper introduces generalized steppingstone sampling (GSS) as a method for estimating marginal likelihoods and Bayes factors in pulsar timing array (PTA) analyses. The authors derive the GSS estimator as an importance-sampling product of ratios along a path from a reference distribution to the posterior, present a Gaussian simulation comparing GSS with thermodynamic integration and steppingstone sampling, and apply GSS to NANOGrav 15-year single-pulsar and common-red-process models and to EPTA+InPTA DR2 datasets. They report that GSS reproduces the NANOGrav GWB evidence with smaller uncertainty than previously published and find, for the EPTA data, log Bayes factors for GWB over CURN between 5.70 and 8.24, which they describe as a substantial increase over the evidence reported in the original EPTA analysis.

Significance. If the PTA results are reliable, the paper would provide a valuable addition to the PTA model-selection toolbox: GSS is mathematically well founded, it avoids the arbitrary model weights of the hypermodel approach, it can compare non-nested and expensive models such as HD and ORF directly, and the authors ship reproducible code and public data. The Gaussian simulations are clean and confirm the estimator's unbiasedness when the reference distribution is well specified. The NANOGrav consistency check is a useful sanity test. However, the headline EPTA claim rests on unvalidated assumptions about the reference distribution and on small effective sample sizes, so the significance of the reported evidence increase is not yet established.

major comments (3)
  1. [§4.1.2, §4.2, Eq. (14)–(16)] The unbiasedness and finite-variance properties of the GSS ratios in Eq. (14) require the reference distribution π0 to have sufficient probability mass over every power posterior p_β along the path, especially in the tails. In the PTA applications π0 is a product of independent Normal distributions calibrated from posterior sample means and variances. PTA posteriors for the common-process parameters such as log10 A and γ are known to be correlated and non-Gaussian, and a product of marginals can under-cover the joint tails. The paper reports only the standard deviation of repeated log z estimates, which does not detect a common bias, and it does not provide diagnostics for the importance weights (e.g., effective sample size of the ratios, largest-weight fraction) or a sensitivity analysis varying π0. The authors' §5 caveat that a poor posterior approximation 'requires more effort' acknowledges the risk without showing that K = 16 and n_ESS ≈ 20 suffice. I ask for explicit weight diagnostics and a reference-distribution sensitivity test before the EPTA evidence increase can be accepted.
  2. [§4.2, Table 5, Abstract] The central claim of a 'substantial increase in evidence supporting GWB' relative to the EPTA second data release is not quantified in the paper. Table 5 gives GSS values of log BF(GWB/CURN), but the original published log Bayes factors from Antoniadis et al. (2023b) for the same datasets are not stated, so the reader cannot verify the size or direction of the claimed increase. Please report the original values explicitly, with the differences and uncertainties, for each of DR2new, DR2new+, DR2full, and DR2full+.
  3. [§3, Figures 1–2] The Gaussian simulation study uses a model whose posterior is exactly a product of independent Normals, so the product-of-Normals reference distribution is correctly specified up to calibration error. This is the ideal case for GSS and does not exercise the misspecification regime that is the main risk in the PTA application. The demonstration that GSS works with n = 10 or n = 50 samples per chain therefore does not by itself justify the same accuracy for correlated, skewed PTA posteriors. I recommend adding a simulation with a correlated or otherwise non-product posterior, or an intentionally misspecified reference distribution, to characterize when the reported estimator remains unbiased and when it fails.
minor comments (5)
  1. [§2.1, Eq. (9)] For the steppingstone definition q_β = L^β π, the ratio should be E_{p_{β_{k-1}}}[L^{β_k−β_{k-1}}], not E_{p_{β_{k-1}}}[(Lπ)^{β_k−β_{k-1}}]; the extra factor of π makes Eq. (9) inconsistent with Eq. (6).
  2. [§2.1, Eq. (12)] The wording 'β_k = k/K 100% quantile of Beta(α,1)' is ambiguous; it should be stated that β_k is the k/K quantile of Beta(α,1), for example.
  3. [§4.2, Tables 4–5] The abbreviation PSRN is used in the table rows but is not defined in the text; please define it (presumably pulsar-specific red noise) at first use.
  4. [§4.2] The text says the posterior sample means and standard deviations were 'examined for consistency' with Antoniadis et al. (2023b), but the comparison is not shown; a supplementary table or a brief statement of the largest discrepancy would help the reader assess the reproducibility.
  5. [§4.1.2] The statement that using π0 as a proposal distribution 'enables us to skip the new burn-in period' is not self-evident, since the target of each chain is p_β rather than π0; please clarify the initialization and convergence checks for the GSS chains.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: GSS is an unbiased importance-sampling estimator validated against an analytic Gaussian marginal likelihood; the posterior-calibrated reference distribution is a computational device, not a fitted proxy for the reported evidence.

full rationale

The paper's derivation chain is self-contained for the claimed methodological result. GSS is an existing estimator (Fan et al. 2011); the paper re-derives the product-of-ratios identity in Eqs. (8)-(16) and validates it on a Gaussian model with known log z = -115.38, where GSS converges to the analytic value (Fig. 1). The application phase does not fit the marginal likelihood: pi0 in Sec. 4.1.2 is calibrated from posterior sample means and variances, but this only defines the reference distribution in the importance ratio (L pi / pi0)^(Delta beta); the evidence estimate still requires MCMC sampling at each power posterior and the ratios are unbiased Monte Carlo estimates. The NANOGrav and EPTA analyses are re-estimations from public data using the same likelihoods; reproducing posterior means and standard deviations for consistency does not make the Bayes factors inputs. The few self-citations (Maturana-Russel et al. 2019; Maturana-Russel 2017) are for standard steppingstone methodology and numerical experience, not load-bearing for the central claim. The Sec. 5 caveat that a poor posterior approximation requires more effort is an acknowledged efficiency limitation, not a circular definition. Therefore no circular step is exhibited.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The method introduces no new physical entities. The free parameters listed here are algorithmic tuning choices and the reference distribution fitted to posterior samples. The main axioms are that the public PTA data products and the Normal product reference distribution are adequate.

free parameters (5)
  • alpha in Beta(alpha, 1) beta-spacing schedule = 0.3
    Chosen from Xie et al. (2011) as optimal for SS and TI; adopted here without re-optimization for GSS or PTA.
  • Number of beta chains K = 8 to 64
    Selected by the authors; results reported at K = 16. The estimator's consistency is checked by increasing K.
  • Effective sample size per beta chain = 10 to 50
    Target nESS per chain used in PTA applications; very low, relies on the reference distribution being close to the target.
  • Calibration sample size Ncal = 20000
    Number of burn-in posterior samples used to estimate the reference distribution parameters.
  • Reference distribution parameters (per-parameter means and variances) = Estimated from posterior samples
    These are fitted values, but they are algorithmic inputs, not physical parameters. They affect efficiency but not the target value if the estimator is unbiased.
assumptions (3)
  • domain assumption The likelihood functions and noise models implemented in ENTERPRISE for the NANOGrav 15yr and EPTA DR2 data are correct and match the original analyses.
    The paper reproduces prior analyses and relies on public data; errors in the original data products would propagate into the evidence estimates.
  • ad hoc to paper The posterior distributions of the PTA model parameters are approximately unimodal and well approximated by a product of independent Normal distributions for the purpose of the reference distribution.
    The authors fit pi0 as a product of normals calibrated to posterior sample means and variances; if the posterior is multimodal or heavy-tailed, the importance ratios may have high variance.
  • domain assumption The Markov chains for each beta value have converged to their target power posteriors despite the claim that burn-in can be skipped.
    The paper does not provide convergence diagnostics for the GSS MCMC chains; the low nESS per chain (10 to 50) makes convergence a nontrivial requirement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generalized Steppingstone Sampling: Efficient marginal likelihood estimation in gravitational wave analysis of Pulsar Timing Array data." pith.science (2026). https://pith.science/paper/XTVVV2YB

@misc{pith2026241114736,
  author       = {Pith},
  title        = {Pith review of: Generalized Steppingstone Sampling: Efficient marginal likelihood estimation in gravitational wave analysis of Pulsar Timing Array data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XTVVV2YB}},
  note         = {Machine review of arXiv:2411.14736}
}
read the original abstract

Globally, Pulsar Timing Array (PTA) experiments have revealed evidence supporting an existing gravitational wave background (GWB) signal in the PTA data set. Apart from acquiring more observations, the sensitivity of PTA experiments can be increased by improving the accuracy of the noise modeling. In PTA data analysis, noise modeling is conducted primarily using Bayesian statistics, relying on the marginal likelihood and the Bayes factor to assess the evidence. We introduce generalized steppingstone (GSS) as an efficient and accurate marginal likelihood estimation method for the PTA-Bayesian framework. This method enables low-cost estimates with high accuracy, especially when comparing expensive models such as the Hellings-Downs (HD) model or the overlap reduction function model (ORF). We demonstrate the efficiency and the accuracy of GSS for model selection and evidence calculation by reevaluating the evidence of previous analyses from the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) 15 yr data set and the European PTA (EPTA) second data release. We find similar evidence for the GWB compared to the one reported by the NANOGrav 15-year data set. Compared to the evidence reported for the EPTA second data release, we find a substantial increase in evidence supporting GWB across all data sets.

Figures

Figures reproduced from arXiv: 2411.14736 by the authors.

Figure 1
Figure 1. Comparison of log-marginal likelihood estimated as a function of 𝐾 number of 𝛽 chains for the Gaussian model with TI, SS, and GSS methods. Thus, only increasing 𝐾 will ensure convergence and an accurate log(𝑧) estimate. The GSS estimator’s performance remains consistent with a higher number of dimensions as shown in [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Convergence of the marginal likelihood estimated using GSS for the Gaussian model with 𝑑 = 2000 and 𝑣 = 0.01. Top panel: Comparison of log-marginal likelihood estimated as a function of 𝐾 number of 𝛽 chains with 𝑛 = 10. Bottom panel: Comparison of log-marginal likelihood estimated as a function of 𝑛 number of independent samples per 𝛽 chains with 𝐾 = 4 . of the reproduced Bayesian analysis is compared with the resul… view at source ↗
Figure 3
Figure 3. Estimations as a function of 𝐾 number of 𝛽 chains for the NANOGrav dataset via GSS. Top panel: log-marginal likelihood estimated for the IRN model. Middle panel: log-marginal likelihood estimated for the CURN model. Bottom panel: log(BFcCURN/IRN ) comparing CURN to IRN. log(b𝑧) std IRN 7353013.52 1.07 CURN 7353042.72 0.84 Spl ORF 7353037.75 1.18 HD 7353047.80 0.91 [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Mean and standard deviation of 1000 independent log(BFcM1 /M2 ) estimates comparing different common red process models in the NANOGrav dataset where the arrow starts at model M2 and ends at model M1. ported ones in Antoniadis et al. (2023b). A burn-in period of the fi…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 7 canonical work pages

  1. [1]

    Afzal A., et al., 2023, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/acdc91 , 951, L11

    doi:10.3847/2041-8213/acdc91 2023

  2. [2]

    Agazie G., et al., 2023, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/acdac6 , 951, L8

    doi:10.3847/2041-8213/acdac6 2023

  3. [3]

    Antoniadis J., et al., 2023a, @doi [Astronomy & Astrophysics] 10.1051/0004-6361/202346842 , 678, A49

    doi:10.1051/0004-6361/202346842

  4. [4]

    Antoniadis J., et al., 2023b, @doi [Astronomy & Astrophysics] 10.1051/0004-6361/202346844 , 678, A50

    doi:10.1051/0004-6361/202346844

  5. [5]

    Arzoumanian Z., et al., 2020, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/abd401 , 905, L34

    doi:10.3847/2041-8213/abd401 2020

  6. [6]

    P., Chib S., 1995, @doi [Journal Of The Royal Statistical Society Series B: Statistical Methodology] 10.1111/j.2517-6161.1995.tb02042.x , 57, 473

    Carlin B. P., Chib S., 1995, @doi [Journal Of The Royal Statistical Society Series B: Statistical Methodology] 10.1111/j.2517-6161.1995.tb02042.x , 57, 473

    arXiv 1995

  7. [7]

    Chen S., et al., 2021, @doi [Monthly Notices of the Royal Astronomical Society] 10.1093/mnras/stab2833 , 508, 4970

    doi:10.1093/mnras/stab2833 2021

  8. [8]

    Chib S., Greenberg E., 1995, @doi [The American Statistician] 10.1080/00031305.1995.10476177 , 49, 327

    arXiv 1995

Show all 50 references
  1. [9]

    M., Shannon R

    Cordes J. M., Shannon R. M., 2010, A Measurement Model for Precision Pulsar Timing ( @eprint arXiv 1010.3785 ), https://arxiv.org/abs/1010.3785

    2010 arXiv

  2. [10]

    Dalc \' n L., Paz R., Storti M., 2005, @doi [Journal of Parallel and Distributed Computing] 10.1016/j.jpdc.2005.03.010 , 65, 1108

    2005 doi

  3. [11]

    Ellis J., van Haasteren R., 2017, jellis18/PTMCMCSampler: Official Release, @doi 10.5281/zenodo.1037579 , https://doi.org/10.5281/zenodo.1037579

    2017 doi

  4. [12]

    A., Vallisneri M., Taylor S

    Ellis J. A., Vallisneri M., Taylor S. R., Baker P. T., 2019, @doi [Astrophysics Source Code Library] 10.5281/zenodo.4059815 , pp ascl--1912

    2019 doi

  5. [13]

    O., 2011, @doi [Molecular biology and evolution] 10.1093/molbev/msq224 , 28, 523

    Fan Y., Wu R., Chen M.-H., Kuo L., Lewis P. O., 2011, @doi [Molecular biology and evolution] 10.1093/molbev/msq224 , 28, 523

    2011 doi

  6. [14]

    D., et al., 2010, @doi [Classical and Quantum Gravity] 10.1088/0264-9381/27/8/084014 , 27, 084014

    Ferdman R. D., et al., 2010, @doi [Classical and Quantum Gravity] 10.1088/0264-9381/27/8/084014 , 27, 084014

    2010 doi

  7. [15]

    N., 2008, @doi [Journal of the Royal Statistical Society Series B: Statistical Methodology] 10.1111/j.1467-9868.2007.00650.x , 70, 589

    Friel N., Pettitt A. N., 2008, @doi [Journal of the Royal Statistical Society Series B: Statistical Methodology] 10.1111/j.1467-9868.2007.00650.x , 70, 589

    2008

  8. [16]

    Gelman A., Meng X.-L., 1998, Statistical Science, pp 163--185

    1998

  9. [17]

    Goncharov B., et al., 2022, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/ac76bb , 932, L22

    2022 doi

  10. [18]

    R., et al., 2020, @doi [Nature] 10.1038/s41586-020-2649-2 , 585, 357

    Harris C. R., et al., 2020, @doi [Nature] 10.1038/s41586-020-2649-2 , 585, 357

    2020 doi

  11. [19]

    S., Romano J

    Hazboun J. S., Romano J. D., Smith T. L., 2019, @doi [Physical Review D] 10.1103/PhysRevD.100.104028 , 100, 104028

    2019 doi

  12. [20]

    265, Feb

    Hellings R., Downs G., 1983, @doi [Astrophysical Journal, Part 2-Letters to the Editor, vol. 265, Feb. 15, 1983, p. L39-L42.] 10.1086/183954 , https://ui.adsabs.harvard.edu/abs/1983ApJ...265L..39H 265, L39

    1983 doi

  13. [21]

    F., van den Bergh D., Wagenmakers E.-J., 2020, @doi [Advances in Methods and Practices in Psychological Science] 10.1177/2515245919898657 , 3, 200

    Hinne M., Gronau Q. F., van den Bergh D., Wagenmakers E.-J., 2020, @doi [Advances in Methods and Practices in Psychological Science] 10.1177/2515245919898657 , 3, 200

    2020 doi

  14. [22]

    Hourihane S., Meyers P., Johnson A., Chatziioannou K., Vallisneri M., 2023, @doi [Physical Review D] 10.1103/PhysRevD.107.084045 , 107, 084045

    2023 doi

  15. [23]

    D., 2007, @doi [Computing in Science & Engineering] 10.1109/MCSE.2007.55 , 9, 90

    Hunter J. D., 2007, @doi [Computing in Science & Engineering] 10.1109/MCSE.2007.55 , 9, 90

    2007 doi

  16. [24]

    D., et al., 2024, @doi [ ] 10.1103/PhysRevD.109.103012 , https://ui.adsabs.harvard.edu/abs/2024PhRvD.109j3012J 109, 103012

    Johnson A. D., et al., 2024, @doi [ ] 10.1103/PhysRevD.109.103012 , https://ui.adsabs.harvard.edu/abs/2024PhRvD.109j3012J 109, 103012

    2024 doi

  17. [25]

    L., et al., 2017, @doi [The Astrophysical Journal] 10.3847/1538-4357/aa73df , 841, 125

    Jones M. L., et al., 2017, @doi [The Astrophysical Journal] 10.3847/1538-4357/aa73df , 841, 125

    2017 doi

  18. [26]

    C., et al., 2018, @doi [Journal of Astrophysics and Astronomy] 10.1007/s12036-018-9549-y , 39, 51

    Joshi B. C., et al., 2018, @doi [Journal of Astrophysics and Astronomy] 10.1007/s12036-018-9549-y , 39, 51

    2018 doi

  19. [27]

    IOS press, pp 87--90, @doi 10.3233/978-1-61499-649-1-87

    Kluyver T., et al., 2016, in , Positioning and power in academic publishing: Players, agents and agendas. IOS press, pp 87--90, @doi 10.3233/978-1-61499-649-1-87

    2016 doi

  20. [28]

    Kumar R., Carroll C., Hartikainen A., Martin O., 2019, @doi [Journal of Open Source Software] 10.21105/joss.01143 , 4, 1143

    2019 doi

  21. [29]

    Lartillot N., Philippe H., 2006, @doi [Systematic Biology] 10.1080/10635150500433722 , 55, 195

    2006 doi

  22. [30]

    P., Feroz F., van Haasteren R., Lee K., Shannon R

    Lentati L., Alexander P., Hobson M. P., Feroz F., van Haasteren R., Lee K., Shannon R. M., 2014, @doi [Monthly Notices of the Royal Astronomical Society] 10.1093/mnras/stt2122 , 437, 3004

    2014 doi

  23. [31]

    D., Tuerlinckx F., Kuppens P., Wagenmakers E.-J., 2011, @doi [Journal of Mathematical Psychology] 10.1016/j.jmp.2011.06.001 , 55, 331

    Lodewyckx T., Kim W., Lee M. D., Tuerlinckx F., Kuppens P., Wagenmakers E.-J., 2011, @doi [Journal of Mathematical Psychology] 10.1016/j.jmp.2011.06.001 , 55, 331

    2011 doi

  24. [32]

    Manchester R., et al., 2013, @doi [Publications of the Astronomical Society of Australia] 10.1017/pasa.2012.017 , 30, e017

    2013 doi

  25. [33]

    N., Taylor J

    Matsakis D. N., Taylor J. H., Eubanks T. M., 1997, Astronomy and Astrophysics, v. 326, p. 924-928, 326, 924

    1997

  26. [34]

    A., 2017, PhD thesis, University of Auckland, https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm

    Maturana-Russel P. A., 2017, PhD thesis, University of Auckland, https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm

    2017

  27. [35]

    Maturana-Russel P., Meyer R., Veitch J., Christensen N., 2019, @doi [Physical Review D] 10.1103/PhysRevD.99.084006 , 99, 084006

    2019 doi

  28. [36]

    A., 2013, @doi [Classical and Quantum Gravity] 10.1088/0264-9381/30/22/224008 , 30, 224008

    McLaughlin M. A., 2013, @doi [Classical and Quantum Gravity] 10.1088/0264-9381/30/22/224008 , 30, 224008

    2013 doi

  29. [37]

    H., 1996, Statistica Sinica, pp 831--860

    Meng X.-L., Wong W. H., 1996, Statistica Sinica, pp 831--860

    1996

  30. [38]

    A., Raftery A

    Newton M. A., Raftery A. E., 1994, @doi [Journal of the Royal Statistical Society Series B: Statistical Methodology] 10.1111/j.2517-6161.1994.tb01956.x , 56, 3

    1994

  31. [39]

    J., et al., 2023, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/acdd02 , 951, L6

    Reardon D. J., et al., 2023, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/acdd02 , 951, L6

    2023 doi

  32. [40]

    V., 1978, , https://ui.adsabs.harvard.edu/abs/1978SvA....22...36S 22, 36

    Sazhin M. V., 1978, , https://ui.adsabs.harvard.edu/abs/1978SvA....22...36S 22, 36

    1978

  33. [41]

    Skilling J., 2006, @doi [Bayesian Analysis] 10.1214/06-BA127 , 1, 833

    2006 doi

  34. [42]

    Smarra C., et al., 2023, @doi [Physical Review Letters] 10.1103/PhysRevLett.131.171001 , 131, 171001

    2023 doi

  35. [43]

    H., Romani d

    Stinebring D., Ryba M., Taylor J. H., Romani d. R., 1990, @doi [Physical Review Letters] 10.1103/PhysRevLett.65.285 , 65, 285

    1990 doi

  36. [44]

    R., Baker P

    Taylor S. R., Baker P. T., Hazboun J. S., Simon J., Vigeland S. J., 2021, enterprise\_extensions, https://github.com/nanograv/enterprise_extensions

    2021

  37. [45]

    Tiburzi C., et al., 2016, @doi [Monthly Notices of The Royal Astronomical Society] 10.1093/mnras/stv2143 , 455, 4339

    2016 doi

  38. [47]

    Verbiest J., et al., 2009, @doi [Monthly Notices of the Royal Astronomical Society] 10.1111/j.1365-2966.2009.15508.x , 400, 951

    2009

  39. [48]

    Virtanen P., et al., 2020, scipy/scipy: SciPy 1.6.0, @doi 10.5281/zenodo.4406806 , https://doi.org/10.5281/zenodo.4406806

    2020 doi

  40. [49]

    O., Fan Y., Kuo L., Chen M.-H., 2011, @doi [Systematic Biology] 10.1093/sysbio/syq085 , 60, 150

    Xie W., Lewis P. O., Fan Y., Kuo L., Chen M.-H., 2011, @doi [Systematic Biology] 10.1093/sysbio/syq085 , 60, 150

    2011 doi

  41. [50]

    van Haasteren R., Vallisneri M., 2014, @doi [Physical Review D] 10.1103/PhysRevD.90.104012 , 90, 104012

    2014 doi

  42. [51]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.stat...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.