Pith. sign in

REVIEW 7 minor 113 references

Sampling whitened Fourier coefficients turns pulsar-timing gravitational-wave inference from a months-long cluster job into a 15-minute GPU run.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-10 20:19 UTC pith:C3ETLIPO

load-bearing objection Practical, GPU-ready whitening of the Fourier-coefficient posterior that turns HMC into a real workhorse for PTA models; the speed claims hold up on real NG15 data.

arxiv 2607.06834 v1 pith:C3ETLIPO submitted 2026-07-07 gr-qc

A new framework for lightning-fast gravitational wave analysis of pulsar timing data

Aiden Gundersen , Rutger van Haasteren , Neil J. Cornish , Michele Vallisneri , Patrick M. Meyers This is my paper
classification gr-qc
keywords pulsar timing arraysgravitational wave backgroundHamiltonian Monte CarloFourier-domain samplingstandardizing transformhierarchical Bayesian inferenceNANOGrav
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Pulsar timing array analyses have been bottlenecked by the cost of evaluating posteriors after the Fourier coefficients of red noise and a gravitational-wave background are analytically integrated out. This paper keeps those coefficients as free parameters and shows that a simple, hyper-parameter-dependent coordinate change makes them nearly independent standard normals. Hamiltonian Monte Carlo then samples the high-dimensional space efficiently, recovering the same hyper-parameter posteriors as the classic marginalization while evaluating only cheap compressed-matrix products. On data comparable to the NANOGrav 15-year set the full joint model (Hellings–Downs background, pulsar red noise, optional continuous-wave signal) converges in roughly fifteen minutes on one consumer GPU. The same transform also admits free spectral models, inter-frequency correlations, and non-Gaussian priors that were previously impractical.

Core claim

A coordinate transformation that recenters and rescales the Fourier coefficients by their approximate conditional mean and Cholesky factor (computed under a common-uncorrelated-red-noise approximation) converts the hierarchical PTA posterior into a near-standard-normal density that Hamiltonian Monte Carlo can sample in minutes, yielding identical inference on spectral hyper-parameters and deterministic signals without ever inverting large dense covariance matrices.

What carries the argument

The standardizing (whitening) transform a = â_CURN + L_CURN z (and its deterministic-signal generalization), which maps the funnel-shaped Fourier-coefficient posterior onto approximately independent standard normals while leaving the full Hellings–Downs correlations inside the target density.

Load-bearing premise

The whitening matrix can safely ignore all inter-pulsar correlations and still leave the transformed variables close enough to standard normal for efficient sampling.

What would settle it

Run the same NANOGrav-15-year-scale model both with the CURN whitening and with an exact Hellings–Downs Cholesky factor; if the whitened chains fail to recover the known hyper-parameter posteriors or require orders-of-magnitude more steps, the approximation is inadequate.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Joint stochastic-plus-deterministic models that previously required months of CPU time become routine single-GPU analyses.
  • Free spectral models for every pulsar and the background can be run without neglecting inter-frequency or inter-pulsar correlations.
  • Non-Gaussian priors on the Fourier coefficients become feasible because no analytic marginalization is required.
  • Dataset size can grow (more pulsars, longer spans) without the quadratic-to-cubic scaling of the classic dense-matrix approach.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same whitening idea should transfer to other hierarchical time-domain problems whose latent processes are approximately stationary (e.g., continuous-wave searches with red-noise confusion).
  • Once the coefficients are cheap to sample, evidence calculations via thermodynamic integration become practical for model selection between Gaussian and non-Gaussian backgrounds.
  • GPU-native PTA pipelines can now treat the full multi-pulsar free-spectrum model as a default rather than an expensive special case.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 7 minor

Summary. The paper presents a reparameterization of the hierarchical PTA posterior that samples Fourier coefficients numerically rather than analytically marginalizing them. By applying a standardizing (whitening) transform based on the Laplace mean and a CURN Cholesky factor of the coefficient covariance, the latent coefficients become approximately standard normal and can be sampled efficiently with HMC/NUTS on a GPU. The authors derive the transform for pure stochastic models, then generalize it to deterministic signals (e.g. continuous waves), inter-frequency correlations, tempered likelihoods, split GWB/RN coefficients, and non-Gaussian priors. On NANOGrav 15-year-scale data they report converged posteriors for HD-correlated GWB + pulsar RN (and joint CW) models in ~15 minutes on a single RTX 3090, more than an order of magnitude faster than an equivalent JAX analytically-marginalized baseline and far faster than ENTERPRISE on a CPU cluster. Validation includes direct posterior comparison with ENTERPRISE on the real 15-year data set and recovery of injected parameters on a 100-pulsar simulation.

Significance. If the reported speedups and posterior fidelity hold, this is a substantial practical advance for PTA gravitational-wave analysis. It removes a long-standing computational bottleneck that has restricted joint stochastic–deterministic modeling and free-spectral analyses of full arrays. The derivation is standard hierarchical reparameterization applied carefully to the Lentati-style Fourier-domain posterior; the empirical validation against ENTERPRISE on real data and against known injections is strong; and the authors ship a public implementation (PROMETHEUS) with reproducible figure code. Enabling HD+RN+CW analyses in tens of minutes on a single GPU, rather than hours to months, would materially expand the model space accessible to the PTA community.

minor comments (7)
  1. Fig. 1 caption: typo “hiearchical” → “hierarchical”.
  2. Sec. VI A: the quoted effective-sample rates (~1.53 vs ~0.12 ESS/s) should state explicitly which parameters they refer to (hyper-parameters η only, or including the high-dimensional z). A short table of ESS for AGWB, γGWB, and a few RN amplitudes would make the speed claim fully transparent.
  3. Sec. IV B (after Eq. 29): a brief forward-looking sentence on when the CURN whitening approximation might degrade (e.g. if off-diagonal HD power becomes comparable to the diagonal, or for strongly anisotropic backgrounds) would help readers judge applicability to future data sets, even though the present NANOGrav-scale results support the approximation.
  4. Sec. VI B / Appendix B: multi-modality of continuous-wave posteriors is acknowledged as future work; a short note on whether the tempered standardizing transform of Appendix B was used (or not) in the Fig. 6 run would clarify the current practical status of joint CW sampling.
  5. Eq. (31) and Eq. (38): the placement of det(LCURN) relative to the prior normalization is correct but dense; a one-line reminder that this is the absolute value of the Jacobian determinant of the map z → a would aid readers less familiar with non-centered parameterizations.
  6. Fig. 3: the polynomial scaling fits are quoted for Np > 60; stating the exact fit range and whether timing includes only the log-density evaluation or a full NUTS leapfrog step would improve reproducibility of the scaling claim.
  7. References: a few arXiv-only citations (e.g. recent non-Gaussian GWB papers) may have journal versions by publication time; a final pass would be useful.

Circularity Check

0 steps flagged

No significant circularity: standardizing transform is a standard Laplace/non-centered reparameterization of the same posterior that is sampled; validation is against ENTERPRISE and injections.

full rationale

The paper's central claim is a computational reparameterization (standardizing / whitening transform of Fourier coefficients under a CURN Laplace approximation of mean and covariance, Eqs. 29–31 and generalizations) that removes Neal's funnel so HMC/NUTS can sample the hyper-efficient Lentati-style posterior numerically rather than analytically marginalizing. The transform is constructed from the target density itself (MAP and Hessian of Eq. 28, then CURN approx for cost), samples are mapped back by the inverse, and the Metropolis step evaluates the full (HD-correlated) density; inference is therefore identical by construction to the untransformed model, not a prediction forced by a fitted input. Empirical speed and correctness are demonstrated by reproducing NANOGrav 15-year power-law and free-spectrum posteriors against ENTERPRISE (Figs. 4–5) and recovering injected CW+HD+RN parameters (Fig. 6). Self-citations supply prior posterior formulations and sampling attempts; none is a load-bearing uniqueness theorem or ansatz that forces the 15-minute claim. The CURN whitening approximation is explicitly approximate and is justified by published Bayes factors on the same data, not by circular definition. No step reduces a claimed first-principles result to its own inputs.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

The central speed claim rests on standard Bayesian hierarchical modeling, the Laplace approximation for local mean/covariance, the empirical observation that CURN dominates HD correlations in current data, and the known efficiency of HMC/NUTS on near-standard-normal targets. No new physical entities are postulated; free parameters are the usual spectral hyper-parameters and the discrete frequency-bin counts chosen by the analyst.

free parameters (2)
  • Nf (number of Fourier frequency bins) = 14 / 30 / 60 (analysis-dependent)
    Chosen by the analyst (Nf=14 for GWB, 30 for RN in the NANOGrav run; 60 for the CW Fourier representation); directly controls both model fidelity and computational cost.
  • hyper-prior bounds on log10 A and gamma
    Uniform ranges (e.g. log10 A ~ U(-20,-10), gamma ~ U(0,7)) are conventional but still free choices that affect the posterior volume.
axioms (4)
  • domain assumption The Laplace approximation supplies a sufficiently accurate local mean and covariance for the Fourier coefficients that the subsequent linear whitening yields near-standard-normal variables.
    Invoked in Sec. IV B to justify the standardizing map; validity is checked only empirically.
  • ad hoc to paper A CURN (diagonal) approximation to the prior covariance is adequate for the whitening transform even though the target posterior retains full Hellings–Downs correlations.
    Explicit modeling choice after Eq. 29; justified by the large Bayes factor favoring CURN over pure noise and the modest BF of HD over CURN in NANOGrav 15-yr data.
  • domain assumption White-noise parameters (EFAC/EQUAD/ECORR) may be fixed from single-pulsar analyses without biasing the multi-pulsar inference.
    Standard PTA practice restated in Sec. III B; not re-validated here.
  • standard math HMC/NUTS with a Euclidean-Gaussian kinetic energy efficiently samples high-dimensional near-standard-normal targets.
    Background fact from the HMC literature (Appendix A).

pith-pipeline@v1.1.0-grok45 · 36551 in / 2673 out tokens · 38473 ms · 2026-07-10T20:19:46.422220+00:00 · methodology

0 comments
read the original abstract

Pulsar timing array data analysis is computationally expensive, limiting the complexity of models which can be studied. As pulsar timing datasets and their respective models grow in size and sophistication, faster and scalable inference methods are essential. In this paper, we accelerate pulsar timing analyses by sampling in the space of Fourier coefficients instead of analytically marginalizing over them. Previous studies have shown the Fourier space induces a complex, high-dimensional posterior geometry, from which it is generally difficult to sample. We show that under an appropriate coordinate transformation the Fourier coefficients approximately follow a standard normal distribution, and may be efficiently sampled using a Hamiltonian Monte Carlo scheme. Under this coordinate transformation, for datasets of size and complexity comparable to the NANOGrav 15-year release, the new method produces converged posterior distributions for a range of models which include inter-pulsar correlations, stochastic, and deterministic signals in approximately 15 minutes on an NVIDIA GeForce RTX 3090 GPU. By comparison, the legacy pulsar timing analysis software \texttt{ENTERPRISE} would require months of computation on a CPU cluster to analyze comparable datasets under the same joint stochastic and deterministic models.

Figures

Figures reproduced from arXiv: 2607.06834 by Aiden Gundersen, Michele Vallisneri, Neil J. Cornish, Patrick M. Meyers, Rutger van Haasteren.

Figure 1
Figure 1. Figure 1: FIG. 1. Samples from the hiearchical toy model, Eq. (5) and [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Frequency-domain prior covariance matrices, [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Evaluation times of the posterior on an NVIDIA [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Corner plot illustrating the recovery of the GWB, J1745+1017’s noise spectrum, and J1853+1303’s noise spectrum [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Free spectral analysis of the stochastic GWB in the NANOGrav 15-year dataset. The violins show the timing delays [PITH_FULL_IMAGE:figures/full_fig_p017_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Corner plot illustrating the recovery of a subset of CW parameters, the 3 [PITH_FULL_IMAGE:figures/full_fig_p018_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. Recovery of a simulated gravitational wave back [PITH_FULL_IMAGE:figures/full_fig_p023_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

113 extracted references · 113 canonical work pages · 22 internal anchors

  1. [1]

    These meth- ods are currently being added to theDISCOVERY[95] andATLAS(in preparation) packages as well for next- generation PTA analyses

    for rapid PTA parameter estimation. These meth- ods are currently being added to theDISCOVERY[95] andATLAS(in preparation) packages as well for next- generation PTA analyses. WithinPROMETHEUS, the pos- terior density and standardizing transform are imple- mented inJAX[17]. The sampling is performed with NUMPYRO[83, 84]. Figures are created withMATPLOTLIB

    work page

  2. [2]

    GWB” and “RN

    andCORNER[97]. The simulated datasets and code to reproduce figures presented in this work are publicly available at [98]. Data. Simulated data is generated with theNUMPY package [99]. The North American Nanohertz Observa- tory for Gravitational Waves (NANOGrav) 15-year Data Set [100] is analyzed using the methods presented in this paper to reproduce exis...

    work page

  3. [3]

    M. V. Sazhin, Sov. Astron.22, 36 (1978)

    work page 1978

  4. [4]

    Detweiler, Astrophys

    S. Detweiler, Astrophys. J.234, 1100 (1979)

    work page 1979

  5. [5]

    Agazieet al.(NANOGrav Collaboration), The As- trophysical Journal Letters951, L8 (2023)

    G. Agazieet al.(NANOGrav Collaboration), The As- trophysical Journal Letters951, L8 (2023)

    work page 2023

  6. [6]

    D. J. Reardonet al.(Parkes Pulsar Timing Array), The Astrophysical Journal Letters951, L6 (2023)

    work page 2023

  7. [7]

    The second data release from the European Pulsar Timing Array III. Search for gravitational wave signals

    EPTA Collaboration, InPTA Collaboration,et al., As- tron. Astrophys.678, A50 (2023), arXiv:2306.16214 [astro-ph.HE]

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  8. [8]

    Searching for the nano-Hertz stochastic gravitational wave background with the Chinese Pulsar Timing Array Data Release I

    H. Xuet al.(Chinese Pulsar Timing Array), Research in Astronomy and Astrophysics23, 075024 (2023), arXiv:2306.16216 [astro-ph.HE]

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  9. [9]

    M. T. Mileset al.(MeerKAT Pulsar Timing Array), Monthly Notices of the Royal Astronomical Society536, 1489 (2025), https://academic.oup.com/mnras/article- pdf/536/2/1489/61215196/stae2571.pdf

    work page 2025

  10. [10]

    Ultra-Low Frequency Gravitational Radiation from Massive Black Hole Binaries

    M. Rajagopal and R. W. Romani, Astrophys. J.446, 543 (1995), arXiv:astro-ph/9412038 [astro-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 1995

  11. [11]

    A. H. Jaffe and D. C. Backer, The Astrophysical Journal 583, 616 (2003)

    work page 2003

  12. [12]

    J. S. B. Wyithe and A. Loeb, The Astrophysical Journal 590, 691 (2003)

    work page 2003

  13. [13]

    Low-frequency gravitational radiation from coalescing massive black hole binaries in hierarchical cosmologies

    A. Sesana, F. Haardt, P. Madau, and M. Volonteri, Astrophys. J.611, 623 (2004), arXiv:astro-ph/0401543 [astro-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2004

  14. [14]

    S. T. McWilliams, J. P. Ostriker, and F. Pretorius, Astrophys. J.789, 156 (2014), arXiv:1211.5377 [astro- ph.CO]

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  15. [15]

    Burke-Spolaor, S

    S. Burke-Spolaor, S. R. Taylor, M. Charisi, T. Dolch, J. S. Hazboun, A. M. Holgado, L. Z. Kelley, T. J. W. Lazio, D. R. Madison, N. McMann, C. M. F. Mingarelli, A. Rasskazov, X. Siemens, J. J. Simon, and T. L. Smith, The Astronomy and Astrophysics Review27, 5 (2019)

    work page 2019

  16. [16]

    Afzalet al.(NANOGrav Collaboration), The Astro- physical Journal Letters951, L11 (2023)

    A. Afzalet al.(NANOGrav Collaboration), The Astro- physical Journal Letters951, L11 (2023)

    work page 2023

  17. [17]

    Vallisneri, M

    M. Vallisneri, M. Crisostomi, A. D. Johnson, and P. M. Meyers, Phys. Rev. Lett.135, 071401 (2025)

    work page 2025

  18. [18]

    D. Shih, M. Freytsis, S. R. Taylor, J. A. Dror, and N. Smyth, Phys. Rev. Lett.133, 011402 (2024)

    work page 2024

  19. [19]

    Bradbury, R

    J. Bradbury, R. Frostig, P. Hawkins, M. J. Johnson, C. Leary, D. Maclaurin, G. Necula, A. Paszke, J. Van- derPlas, S. Wanderman-Milne, and Q. Zhang, JAX: composable transformations of Python+NumPy pro- grams,http://github.com/jax-ml/jax(2018). 24

    work page 2018

  20. [20]

    J. A. Ellis, M. Vallisneri, S. R. Taylor, and P. T. Baker, ENTERPRISE: Enhanced Numerical Toolbox Enabling a Robust PulsaR Inference SuitE, Astrophysics Source Code Library, record ascl:1912.015 (2019)

    work page 1912

  21. [21]

    Lentati, P

    L. Lentati, P. Alexander, M. P. Hobson, S. Taylor, J. Gair, S. T. Balan, and R. van Haasteren, Phys. Rev. D87, 104021 (2013)

    work page 2013

  22. [22]

    van Haasteren and M

    R. van Haasteren and M. Vallisneri, Phys. Rev. D90, 104012 (2014)

    work page 2014

  23. [23]

    N. Laal, W. G. Lamb, J. D. Romano, X. Siemens, S. R. Taylor, and R. van Haasteren, Phys. Rev. D108, 063008 (2023)

    work page 2023

  24. [24]

    N. Laal, S. R. Taylor, R. van Haasteren, W. G. Lamb, and X. Siemens, Phys. Rev. D111, 063067 (2025)

    work page 2025

  25. [25]

    Gundersen and N

    A. Gundersen and N. J. Cornish, Phys. Rev. D112, 083035 (2025)

    work page 2025

  26. [26]

    Gundersen and N

    A. Gundersen and N. J. Cornish, arXiv e-prints (2025), arXiv:2510.12917 [stat.ME]

    work page arXiv 2025

  27. [27]

    S. R. Taylor, J. Simon, and L. Sampson, Phys. Rev. Lett.118, 181102 (2017)

    work page 2017

  28. [28]

    Vallisneri and R

    M. Vallisneri and R. van Haasteren, Monthly Notices of the Royal Astronomical Society466, 4954 (2017)

    work page 2017

  29. [29]

    A Conceptual Introduction to Hamiltonian Monte Carlo

    M. Betancourt, arXiv e-prints (2017), arXiv:1701.02434 [stat.ME]

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  30. [30]

    Gelman, W

    A. Gelman, W. R. Gilks, and G. O. Roberts, The Annals of Applied Probability7, 110 (1997)

    work page 1997

  31. [31]

    Beskos, N

    A. Beskos, N. Pillai, G. Roberts, J.-M. Sanz-Serna, and A. Stuart, Bernoulli19, 1501 (2013)

    work page 2013

  32. [32]

    G. E. Freedman, A. D. Johnson, R. van Haasteren, and S. J. Vigeland, Phys. Rev. D107, 043013 (2023)

    work page 2023

  33. [33]

    G. E. Freedman and S. J. Vigeland, Phys. Rev. D110, 063038 (2024)

    work page 2024

  34. [34]

    Buikemaet al.(LIGO Scientific Collaboration), Phys

    A. Buikemaet al.(LIGO Scientific Collaboration), Phys. Rev. D102, 062003 (2020)

    work page 2020

  35. [35]

    Acerneseet al.(Virgo Collaboration), Phys

    F. Acerneseet al.(Virgo Collaboration), Phys. Rev. Lett.123, 231108 (2019)

    work page 2019

  36. [36]

    A. G. Abacet al.(LIGO Scientific, Virgo, and KAGRA Collaborations), Gwtc-5.0: An introduction to version 5.0 of the gravitational-wave transient catalog (2026), arXiv:2605.27223 [gr-qc]

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  37. [37]

    M. R. Adams, N. J. Cornish, and T. B. Littenberg, Phys. Rev. D86, 124032 (2012), arXiv:1209.6286 [gr- qc]

    work page internal anchor Pith review Pith/arXiv arXiv 2012

  38. [38]

    Mandel, W

    I. Mandel, W. M. Farr, and J. R. Gair, Monthly No- tices of the Royal Astronomical Society486, 1086–1093 (2019)

    work page 2019

  39. [39]

    Vitale, D

    S. Vitale, D. Gerosa, W. M. Farr, and S. R. Taylor, Inferring the properties of a population of compact bi- naries in presence of selection effects, inHandbook of Gravitational Wave Astronomy, edited by C. Bambi, S. Katsanevas, and K. D. Kokkotas (Springer Singapore, Singapore, 2020) pp. 1–60

    work page 2020

  40. [40]

    Mancarella and D

    M. Mancarella and D. Gerosa, Phys. Rev. D111, 103012 (2025)

    work page 2025

  41. [41]

    S. R. Taylor and D. Gerosa, Phys. Rev. D98, 083017 (2018)

    work page 2018

  42. [42]

    van Haasteren, The Astrophysical Journal Supple- ment Series273, 23 (2024)

    R. van Haasteren, The Astrophysical Journal Supple- ment Series273, 23 (2024)

    work page 2024

  43. [43]

    Goncharov, E

    B. Goncharov, E. Thrane, R. M. Shannon, J. Harms, N. D. R. Bhat, G. Hobbs, M. Kerr, R. N. Manchester, D. J. Reardon, C. J. Russell, X.-J. Zhu, and A. Zic, The Astrophysical Journal Letters932, L22 (2022)

    work page 2022

  44. [44]

    Goncharov and S

    B. Goncharov and S. Sardana, Monthly Notices of the Royal Astronomical Society537, 3470 (2025), https://academic.oup.com/mnras/article- pdf/537/4/3470/61743825/staf190.pdf

    work page 2025

  45. [45]

    Gelman, J

    A. Gelman, J. B. Carlin, H. S. Stern, D. B. Dunson, A. Vehtari, and D. B. Rubin,Bayesian Data Analysis (Chapman and Hall/CRC, 2013)

    work page 2013

  46. [46]

    R. M. Neal, The Annals of Statistics31, 705 (2003)

    work page 2003

  47. [47]

    Metropolis, A

    N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller, J. Chem. Phys.21, 1087 (1953)

    work page 1953

  48. [48]

    Girolami and B

    M. Girolami and B. Calderhead, Journal of the Royal Statistical Society: Series B (Statistical Methodology)73, 123 (2011), https://rss.onlinelibrary.wiley.com/doi/pdf/10.1111/j.1467- 9868.2010.00765.x

    work page doi:10.1111/j.1467- 2011

  49. [49]

    Betancourt, inGeometric Science of Information, edited by F

    M. Betancourt, inGeometric Science of Information, edited by F. Nielsen and F. Barbaresco (Springer Berlin Heidelberg, Berlin, Heidelberg, 2013) pp. 327–334

    work page 2013

  50. [50]

    Papaspiliopoulos, G

    O. Papaspiliopoulos, G. O. Roberts, and M. Sk¨ old, Sta- tistical Science22, 59 (2007)

    work page 2007

  51. [52]

    Agazieet al.(NANOGrav Collaboration), The As- trophysical Journal Letters951, L10 (2023)

    G. Agazieet al.(NANOGrav Collaboration), The As- trophysical Journal Letters951, L10 (2023)

    work page 2023

  52. [53]

    R. M. Shannon and J. M. Cordes, The Astrophysical Journal725, 1607 (2010)

    work page 2010

  53. [54]

    E. S. Phinney, arXiv e-prints (2001), arXiv:astro- ph/0108028

    work page arXiv 2001

  54. [55]

    M. J. Keith, W. Coles, R. M. Shannon, G. B. Hobbs, R. N. Manchester, M. Bailes, N. D. R. Bhat, S. Burke- Spolaor, D. J. Champion, A. Chaudhary, A. W. Hotan, J. Khoo, J. Kocz, S. Os lowski, V. Ravi, J. E. Reynolds, J. Sarkissian, W. van Straten, and D. R. B. Yard- ley, Mon. Not. R. Astron. Soc.429, 2161 (2013), arXiv:1211.5887 [astro-ph.GA]

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  55. [56]

    K. J. Lee, C. G. Bassa, G. H. Janssen, R. Karuppusamy, M. Kramer, K. Liu, D. Perrodin, R. Smits, B. W. Stap- pers, R. van Haasteren, and L. Lentati, Monthly No- tices of the Royal Astronomical Society441, 2831–2844 (2014)

    work page 2014

  56. [57]

    Pulsar Timing Array Observations of Massive Black Hole Binaries

    V. Corbin and N. J. Cornish, arXiv e-prints (2010), arXiv:1008.1782 [astro-ph.HE]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  57. [58]

    K. J. Lee, N. Wex, M. Kramer, B. W. Stappers, C. G. Bassa, G. H. Janssen, R. Karuppusamy, and R. Smits, Monthly Notices of the Royal Astronomical Society414, 3251–3264 (2011)

    work page 2011

  58. [59]

    Nonlinear gravitational-wave memory from binary black hole mergers

    M. Favata, Astrophys. J. Lett.696, L159 (2009), arXiv:0902.3660 [astro-ph.SR]

    work page internal anchor Pith review Pith/arXiv arXiv 2009

  59. [60]

    Favata, Phys

    M. Favata, Phys. Rev. D80, 024002 (2009)

    work page 2009

  60. [61]

    D. J. Champion, G. B. Hobbs, R. N. Manchester, R. T. Edwards, D. C. Backer, M. Bailes, N. D. R. Bhat, S. Burke-Spolaor, W. Coles, P. B. Demorest, R. D. Fer- dman, W. M. Folkner, A. W. Hotan, M. Kramer, A. N. Lommen, D. J. Nice, M. B. Purver, J. M. Sarkissian, I. H. Stairs, W. van Straten, J. P. W. Verbiest, and D. R. B. Yardley, The Astrophysical Journal ...

    work page 2010

  61. [62]

    Vallisneri, S

    M. Vallisneri, S. R. Taylor, J. Simon, W. M. Folkner, R. S. Park, C. Cutler, J. A. Ellis, T. J. W. Lazio, S. J. Vigeland, K. Aggarwal, Z. Arzoumanian, P. T. Baker, A. Brazier, P. R. Brook, S. Burke-Spolaor, S. Chatter- jee, J. M. Cordes, N. J. Cornish, F. Crawford, H. T. Cromartie, K. Crowter, M. DeCesar, P. B. Demorest, T. Dolch, R. D. Ferdman, E. C. Fer...

    work page 2020

  62. [63]

    J. A. Ellis, X. Siemens, and J. D. E. Creighton, The Astrophysical Journal756, 175 (2012)

    work page 2012

  63. [64]

    J. A. Ellis, Classical and Quantum Gravity30, 224004 (2013)

    work page 2013

  64. [65]

    Taylor, J

    S. Taylor, J. Ellis, and J. Gair, Phys. Rev. D90, 104028 (2014)

    work page 2014

  65. [66]

    R. W. Hellings and G. S. Downs, Astrophys. J. Lett. 265, L39 (1983)

    work page 1983

  66. [67]

    Sampson, N

    L. Sampson, N. J. Cornish, and S. T. McWilliams, Phys. Rev. D91, 084055 (2015)

    work page 2015

  67. [68]

    Sesana, A

    A. Sesana, A. Vecchio, and C. N. Colacino, Monthly Notices of the Royal Astronomical Society390, 192–209 (2008)

    work page 2008

  68. [69]

    The NANOGrav 15-year Data Set: Constraints on Supermassive Black Hole Binaries from the Gravitational Wave Background

    G. Agazie, et al, and Nanograv Collaboration, Astro- phys. J. Lett.952, L37 (2023), arXiv:2306.16220 [astro- ph.HE]

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  69. [70]

    S. R. Taylor, J. R. Gair, and L. Lentati, Phys. Rev. D 87, 044035 (2013)

    work page 2013

  70. [71]

    S. R. Taylor, J. Simon, L. Schult, N. Pol, and W. G. Lamb, Phys. Rev. D105, 084049 (2022), arXiv:2202.08293 [gr-qc]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  71. [72]

    W. G. Lamb, S. R. Taylor, and R. van Haasteren, Phys. Rev. D108, 103019 (2023)

    work page 2023

  72. [73]

    W. G. Lamb and S. R. Taylor, The Astrophysical Jour- nal Letters971, L10 (2024)

    work page 2024

  73. [74]

    W. G. Lamb, J. M. Wachter, A. Mitridate, S. C. Sarde- sai, B. B´ ecsy, E. L. Hagen, S. R. Taylor, and L. Z. Kelley, Phys. Rev. D113, 123065 (2026)

    work page 2026

  74. [75]

    Kuntz, C

    A. Kuntz, C. Smarra, and M. Vaglio, Looking for non- gaussianity in pulsar timing arrays through the four point correlator (2026), arXiv:2603.12311 [gr-qc]

    work page arXiv 2026

  75. [76]

    The distribution of the gravitational-wave background from supermassive black holes

    G. Sato-Polito and M. Zaldarriaga, The distribution of the gravitational-wave background from supermassive black holes (2024), arXiv:2406.17010 [astro-ph.CO]

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  76. [77]

    X. Xue, Z. Pan, and L. Dai, Non-gaussian statis- tics of nanohertz stochastic gravitational waves (2025), arXiv:2409.19516 [astro-ph.CO]

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  77. [78]

    The Heavy Tailed Non-Gaussianity of the Supermassive Black Hole Gravitational Wave Background

    J. Raidal, J. Urrutia, V. Vaskonen, and H. Veerm¨ ae, The heavy tailed non-gaussianity of the supermas- sive black hole gravitational wave background (2026), arXiv:2604.08506 [astro-ph.CO]

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  78. [79]

    On measuring the gravitational-wave background using Pulsar Timing Arrays

    R. van Haasteren, Y. Levin, P. McDonald, and T. Lu, Mon. Not. R. Astron. Soc.395, 1005 (2009), arXiv:0809.0791 [astro-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2009

  79. [80]

    van Haasteren, Monthly Notices of the Royal Astro- nomical Society429, 55–62 (2012)

    R. van Haasteren, Monthly Notices of the Royal Astro- nomical Society429, 55–62 (2012)

    work page 2012

  80. [81]

    Understanding and analysing time-correlated stochastic signals in pulsar timing

    R. van Haasteren and Y. Levin, Mon. Not. Roy. Astron. Soc.428, 1147 (2013), arXiv:1202.5932 [astro-ph.IM]

    work page internal anchor Pith review Pith/arXiv arXiv 2013

Showing first 80 references.