Pith. sign in

REVIEW 2 major objections 3 minor 68 references

Critical Point Processes Obtained from a Gaussian Random Field with a View Towards Statistics

T0 review · 2 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Critical point processes formed by Gaussian random fields are tractable spatial statistics: the paper derives explicit correlations at any distance and a multivariate central limit theorem for intensity and modified K-function estimators.

desk verdict Solid new theory for critical point processes, but the K-function estimator in Theorem 8 is mis-normalized as written. read the letter →

arxiv 2507.04753 v1 pith:WEXFQ4J5 submitted 2025-07-07 math.PR math.STstat.TH

Julien Chevallier (SVH , LJK) , Jean-François Coeurjolly (LJK , SVH) , Rasmus Waagepetersen This is my paper · ORCID
classification math.PRmath.STstat.TH MSC 60G6060D0562M30
keywords criticalpointprocessGaussianrandomfieldpaircorrelationfunctionRipley'sK-functioncentrallimittheoremKac-RiceformulamultipleWiener-Itôintegralsspatialstatistics
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's goal is to turn the set of critical points—minima, maxima, and saddles—of a smooth stationary isotropic Gaussian random field into a usable spatial point process model. It derives an explicit formula for the intensity and for the cross pair correlation function at any distance, not just asymptotically at zero, so that the attraction or repulsion between critical points can be quantified in full. It proves that two approximate simulation schemes converge in distribution to the exact critical point process. Finally, under increasing-domain asymptotics, it establishes a multivariate central limit theorem for the natural estimator of the intensity together with estimators of a modified Ripley's K-function at several distances. If the results hold, this new point process family becomes amenable to the standard statistical toolkit: estimation, simulation, and joint inference.

What carries the argument

The engine is a chain of Gaussian analytic tools. The Kac-Rice formula converts expected counts of critical points into an integral of a conditional expectation of Hessian determinants at the zero-gradient event, yielding Theorem 2 and Theorem 3. Proposition 2 rewrites centered linear and bilinear statistics as Hermite chaos expansions with coefficients da(r) that depend only on the field, not on the test functions; Lemma 1 bounds the resulting covariance kernels using an integrable envelope Ξ, condition (C.4). The multivariate CLT in Theorem 8 is obtained by representing each chaos component as a multiple Wiener-Itô integral and applying the contraction criterion of Peccati and Taqqu. Condition (C.5), the non-degeneracy of the gradient-plus-Hessian vector at two points separated by r in [η,R], is what makes the bilinear expansion and covariance bounds valid.

What would settle it

Directly compute the determinant of the covariance matrix of (X'(0), X''(0), X'(r e1), X''(r e1)) for a Matérn field at a grid of r values with fixed ν and φ; a zero in (η,R) would violate (C.5). Separately, simulate the pair correlation function g_{0,d}(r) for local extrema in dimension 2 by counting critical points in large windows and compare with the Monte-Carlo evaluation of the Theorem 2 formula; a persistent mismatch beyond sampling error would disprove the explicit expression.

Watch

Extended reading notes

Core claim

Theorem 2(ii) gives, for any distance r where the joint gradient vector {X'(0), X'(r e1)} is non-degenerate, the cross pair correlation function g_{L,L'}(r) = (ρ_L ρ_L')^{-1} f_{V(r)}(0,0) E[|det X''(0)||det X''(r e1)| ι_L{X''(0)} ι_L'{X''(r e1)} | X'(0)=X'(r e1)=0], with the density f_{V(r)}(0,0) expressed in closed form from the correlation function. This extends earlier small-distance results to all distances. Theorem 8(ii) then proves that the vector of centered, scaled estimators of the intensity and of a modified K-function at finitely many distances converges in distribution to a centered Gaussian vector with an explicit covariance matrix, under conditions (C.1)[5], (C.2)[4], (C.4), and (C.5). The modification of Ripley's K-function by an annulus around the origin is a deliberate technical device that keeps Hermite coefficients uniformly bounded.

Load-bearing premise

The load-bearing premise is condition (C.5), that for every distance r in the range [η,R] the Gaussian vector formed by the gradient and Hessian at 0 and at r e1 is non-degenerate; the paper checks this only numerically for the Matérn and random wave models, and if it failed the bilinear machinery and the central limit theorem would not hold as stated.

Editorial extensions

If this is right

  • Minimum contrast estimation of latent-field parameters (e.g., Matérn smoothness ν and scale φ) becomes feasible by matching the explicit intensity and modified K-function.
  • The simulation theorems justify using smoothed lattice fields or spectral averaging as faithful generators of critical point patterns for practical use.
  • The joint CLT yields asymptotic confidence ellipsoids for (ρ̂_L, K̂_η,L(r_1), ..., K̂_η,L(r_m)), with explicit covariance; the same holds for cross-type point processes Y_L, Y_L'.
  • The Hermite expansions and variance formulas transfer to any stationary point process whose pair correlation function satisfies g−1 ∈ L^1, as noted for the linear-statistic variance.
  • The paper's dimension-dependent repulsion findings give concrete guidance on which index sets and which latent fields produce clustered versus repulsive patterns.

Reading between the lines

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

  • A natural next step, not taken in the paper, is to replace condition (C.4) by an oscillatory-integral version so that the random wave model, whose correlations decay too slowly, also falls under the CLT.
  • Because Theorem 2 expresses the pair correlation as a Gaussian expectation under the two-point conditioning, a deterministic quadrature or closed-form evaluation may be possible, making g(r) computable without Monte Carlo error.
  • The η-annulus in the modified K-function is a scaffolding; a plausible conjecture is that the same CLT holds as η → 0 with an additional variance term, so the standard K-function would be covered by a separate limiting argument.
  • The explicit correlation structure could serve as a test bed for detecting non-Gaussianity of an observed point pattern: a fitted critical point process provides a likelihood-free benchmark against which summary statistics of real patterns can be compared.
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

2 major / 3 minor

Summary. The paper studies critical point processes Y_L obtained as critical points of a stationary isotropic Gaussian random field. It gives Kac-Rice based formulas for the intensity and pair correlation at finite distances (Theorem 2), higher-order intensities (Theorem 3), proves convergence of two approximate simulation strategies (Theorems 5-6), and develops an increasing-domain asymptotic theory for linear and bilinear statistics: Hermite expansions (Proposition 2), variance/covariance limits (Theorem 7) and a multivariate CLT (Theorem 8) for the intensity estimator and a modified Ripley K-function estimator. The paper is largely self-contained with detailed appendices and discloses limitations (Section 6) about positivity of variances, the eta>0 truncation, and simulation step (ii).

Significance. If the technical issues are resolved, this is a substantial contribution: Theorem 2(ii) provides the first finite-distance expression for the pair correlation of critical point processes, and Theorems 7-8 give a multivariate asymptotic distribution theory for summary statistics of a new point process class. The proofs are detailed and use credible machinery (Kac-Rice, Hermite chaos expansions, multiple Wiener-Itô integrals), and the conditions are explicitly stated; the numerical verification of (C.5) and the disclosed open problems are honest. However, the central statistical result as currently written contains a normalization inconsistency in the definition of the K-function estimator, so the main CLT is not yet stated correctly.

major comments (2)
  1. [§5.1, Eq. (26); Appendix K.3, Eq. (90)] There is a load-bearing inconsistency in the definition of the modified Ripley's K-function estimator. In (24), Phi_{2,n} contains the factor 1/rho_L^2, and the text correctly states E(Phi_{2,n}) = K_{eta,L}(r). Equation (26) defines hat K_{eta,L}(r) = Phi_{2,n}/(rho_L^2 hat rho_L^2), whose probability limit is K_{eta,L}(r)/rho_L^4, contradicting Theorem 8(ii). Appendix K.3 expands hat K_{eta,L}(r_j)-K_{eta,L}(r_j) as alpha_n * centered Phi_{2,n} + beta_{n,j} * centered Phi_{1,n} with alpha_n = hat rho_L^2/rho_L^2 and beta_{n,j} = K_{eta,L}(r_j)(hat rho_L+rho_L)/rho_L^2; this corresponds to hat K_{eta,L} = (hat rho_L^2/rho_L^2) Phi_{2,n}, which is neither (26) nor the standard plug-in estimator (rho_L^2/hat rho_L^2) Phi_{2,n}, and the sign of the beta term is opposite to the plug-in first-order correction. The asymptotic covariance matrix in Theorem 8(ii) depends on this first-order term, so the proof and the statement do not currently refer to the same estimator. Please correct the definition and re-derive the delta-method term accordingly.
  2. [§5.2, Table 3; Proposition 2(ii), Lemma 1, Theorems 7-8] Condition (C.5) is load-bearing for the bilinear theory and is verified only numerically for the Matérn and random wave models. Since (C.5) enters Proposition 2(ii), Lemma 1, and hence Theorem 8, the applicability of the main theorems to these models is conditional on a numerical check. Please either provide an analytic proof of (C.5) for the models studied (or at least for the Matérn model), or state explicitly in the statements of Theorems 7-8 that the numerical verification in Table 3 is part of the hypothesis for those models, and specify the grid of distances used in the numerical check.
minor comments (3)
  1. [Appendix K.3, paragraph before Eq. (90)] The phrase 'billinear statistics' contains a typo and should read 'bilinear statistics'.
  2. [§5.4, Theorem 8(ii)] The phrase 'there exist a symmetric nonnegative definite matrix' should read 'there exists a symmetric nonnegative definite matrix'.
  3. [§3.4 and Appendix B] For d >= 3 and L = {0,...,d}, the pair correlation function diverges at zero; the informal wording that 'smaller gL(r) means more repulsion' should distinguish this divergence from the repulsive behavior at positive distances.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's derivations rest on Kac-Rice formulas, Hermite chaos expansions, and external CLT results, not on fitted constants or self-citations.

full rationale

The derivation chain is self-contained and benchmarked externally. Theorem 2 derives the pair correlation function via Kac-Rice formulas (Appendix C) using the external references [8, Theorems 6.2 and 6.4] and [7, Lemma 8]; no fitted quantity enters the expression for g_{L,L'}(r). The Hermite expansions in Proposition 2 follow the framework of [32] and [6], and Theorems 7-8 are proved through Gaussian chaos expansions, Mehler's formula, variance bounds under conditions (C.4) and (C.5), and the external CLT result [52, Theorem 11.8.1]. The estimator in (26) is not a fitted parameter renamed as a prediction: Theorem 8(ii) is a distributional claim obtained by applying the multivariate delta method to the joint CLT in Theorem 8(i), with the limiting covariance matrix derived from explicit Hermite coefficients and covariance functions. The author self-citations [21,22,24,66] appear only in introductory examples, background remarks, and the concluding discussion of future minimum-contrast estimation; they are not load-bearing for any theorem. The apparent normalization mismatch between (26) and the expansion in Appendix K.3 flagged by a skeptical reader is an internal-consistency or correctness concern, not a circularity: it does not make any prediction equivalent to its input by construction. Therefore the circularity burden is negligible, and the score is 0.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The central derivations rest on imported results, all stated and credited: the Kac-Rice formula, Gaussian conditioning structure for Hessians given zero gradient ([7, Lemma 8]), moment finiteness and Geman-type approximation ([32,33]), and the Peccati-Taqqu theorem for the CLT. No new physical or mathematical entities are postulated. The only hand-chosen numeric is the scale phi used to calibrate plots to rho_L = 100.

free parameters (1)
  • Scale parameter phi for numerical illustrations = chosen so that rho_L = 100
    In Figures 2, 4, and 5 the Matérn and random wave model scale parameter phi is chosen by hand so that the intensity equals 100. This is calibration of the simulation study, not a fitted constant in the theoretical results.
assumptions (6)
  • standard math Kac-Rice formula (Theorems 9-10, quoted from Azaïs and Wschebor [8])
    Provides the expected value of counting functionals over critical points used to derive Theorems 1-3.
  • domain assumption Regularity theory linking covariance regularity to sample path regularity (Appendix A diagram, relying on [26, Theorem 7] and Tauberian results from [55])
    Converts conditions on c1 and c2 into the almost sure differentiability conditions (C.1) used throughout the paper.
  • domain assumption Conditional Hessian structure given zero gradient ([7, Lemma 8]): the conditional distribution of Hessians at two points given both gradients vanish is zero-mean Gaussian with covariance depending only on c'_2 and c''_2
    Load-bearing for the explicit pair correlation formula in Theorem 2(ii) and for its Monte Carlo evaluation; imported from prior literature.
  • domain assumption Moment finiteness for critical point counts ([33, Theorem 1.2]) and Geman's condition for second moments following [32] and [6]
    Provides E[N(Delta)^p] < infinity and the Lp approximation (27) on which Proposition 1 and the chaos expansions rest.
  • standard math Peccati-Taqqu theorem for multiple Wiener-Itô integrals ([52, Theorem 11.8.1], restated as Theorem 11)
    Gives the central limit theorem for the chaos components used in Theorem 8.
  • standard math Hermite expansion and Mehler's formula machinery for Gaussian functionals of the gradient-Hessian vector
    Underlies Proposition 2 and follows the approach of Estrade and León [32] and Azaïs, Dalmao, and Delmas [6].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Critical Point Processes Obtained from a Gaussian Random Field with a View Towards Statistics." pith.science (2026). https://pith.science/paper/WEXFQ4J5

@misc{pith2026250704753,
  author       = {Pith},
  title        = {Pith review of: Critical Point Processes Obtained from a Gaussian Random Field with a View Towards Statistics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WEXFQ4J5}},
  note         = {Machine review of arXiv:2507.04753}
}
read the original abstract

This paper establishes the theoretical foundation for statistical applications of an intriguing new type of spatial point processes called critical point processes. These point processes, residing in Euclidean space, consist of the critical points of latent smooth Gaussian random fields or of subsets of critical points like minima, saddle points etc. Despite of the simplicity of their definition, the mathematical analysis of critical point processes is non-trivial involving for example deep results on the geometry of random fields, Sobolev space theory, chaos expansions, and multiple Wiener-It{\^o} integrals. We provide explicit expressions for fundamental moment characteristics used in spatial point process statistics like the intensity parameter, the pair correlation function, and higher order intensity functions. The crucial dependence structure (attraction or repulsiveness) of a critical point process is discussed in depth and is in particular related to the dimension of the points and the type of critical points (extrema, saddle points, or all of the critical points). We propose simulation strategies based on spectral methods or smoothing of grid-based simulations and show that resulting approximate critical point process simulations asymptotically converge to the exact critical point process distribution. Finally, under the increasing domain framework, we obtain asymptotic results for linear and bilinear statistics of a critical point process. In particular, we obtain a multivariate central limit theorem for the intensity parameter estimate and a modified version of Ripley's K-function.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

68 extracted references · 66 canonical work pages

  1. [1]

    and Stegun, I

    Abramowitz, M. and Stegun, I. A. (1968). Handbook of mathematical functions with formulas, graphs, and mathematical tables55. US Government printing office

    work page 1968

  2. [2]

    Adams, R. A. and Fournier, J. J. F. (2003). Sobolev spaces, second ed. Pure and Applied Mathematics (Amsterdam)140.Elsevier/AcademicPress,Amsterdam. MR2424078 (2009e:46025)

    work page 2003

  3. [3]

    Adler, R. J. and Taylor, J. E. (2007). Random fields and geometry80. Springer

    work page 2007

  4. [4]

    Arcones, M. A. (1994). Limit theorems for nonlinear functionals of a stationary Gaussian sequence of vectors.The Annals of Probability2242–2274

    work page 1994

  5. [5]

    Armentano, D., Azaïs, J. M. , Dalmao, F., León, J. R. and Mordecki, E. (2023). On the finiteness of the moments of the measure of level sets of random fields.Brazilian Journal of Probability and Statistics37 219–245

    work page 2023

  6. [6]

    Multivariable CLT for critical points

    Azaïs, J.-M. , Dalmao, F. and Delmas, C. (2024). Multivariate CLT for critical points. arXiv:2401.09117

    work page Pith review arXiv 2024

  7. [7]

    and Delmas, C

    Azaïs, J.-M. and Delmas, C. (2022). Mean number and correlation function of critical points of isotropic Gaussian fields and some results on GOE random matrices.Stochastic Processes and their Applications150 411–445

    work page 2022

  8. [8]

    and Wschebor, M

    Azaïs, J.-M. and Wschebor, M. (2009). Level sets and extrema of random processes and fields. John Wiley & Sons

    work page 2009

Show all 68 references
  1. [9]

    , Rubak, E

    Baddeley, A. , Rubak, E. and Turner, R. (2016). Spatial point patterns: methodology and applications with R1. CRC press Boca Raton

    2016

  2. [10]

    Bardeen, J. M. , Bond, J. , Kaiser, N. and Szalay, A. (1986). The statistics of peaks of Gaussian random fields.Astrophysical Journal, Part 1 (ISSN 0004-637X), vol. 304, May 1, 1986, p. 15-61. SERC-supported research.304 15–61

    1986

  3. [11]

    and Hardy, A

    Bardenet, R. and Hardy, A. (2020). Monte Carlo with determinantal point processes.The Annals of Applied Probability30 368-417

    2020

  4. [12]

    and Wichmann, F

    Barthelmé, S., Trukenbrod, H., Engbert, R. and Wichmann, F. (2013). Modeling fixation locations using spatial point processes.Journal of vision13 1–1

    2013

  5. [13]

    and Gayet, D

    Beffara, V. and Gayet, D. (2017). Percolation of random nodal lines. Publications mathématiques de l’IHÉS126 131–176

    2017

  6. [14]

    and Wigman, I

    Beliaev, D., Cammarota, V. and Wigman, I. (2020). No repulsion between critical points for planar Gaussian random fields.Electronic Communications in Probability25 1–13

    2020

  7. [15]

    Berry, M. V. (1977). Regular and irregular semiclassical wavefunctions.Journal of Physics A: Mathematical and General10 2083

    1977

  8. [16]

    Berry, M. V. (2002). Statistics of nodal lines and points in chaotic quantum billiards: perimeter corrections, fluctuations, curvature.Journal of Physics A: Mathematical and General35 3025

    2002

  9. [17]

    and León, J

    Berzin, C., Latour, A. and León, J. (2022). Kac-Rice formula: a contemporary overview of the main results and applications.arXiv preprint arXiv:2205.08742

    2022

  10. [18]

    Biscio, C. A. N. and Lavancier, F. (2016). Quantifying repulsiveness of determinantal point processes. Bernoulli 22 2001-2028

    2016

  11. [19]

    and Sarnak, P

    Canzani, Y. and Sarnak, P. (2019). Topology and nesting of the zero set components of monochromatic random waves.Communications on Pure and Applied Mathematics72 343– 374

    2019

  12. [20]

    and Schwartzman, A

    Cheng, D. and Schwartzman, A. (2018). Expected number and height distribution of critical points of smooth isotropic Gaussian random fields.Bernoulli 24 3422. CRITICAL POINT PROCESSES 55

    2018

  13. [21]

    and W aagepetersen, R.(2020)

    Choiruddin, A., Cuevas-Pacheco, F., Coeurjolly, J.-F. and W aagepetersen, R.(2020). Regularized estimation for highly multivariate log Gaussian Cox processes.Statistics and Computing 30 649–662

    2020

  14. [22]

    Coeurjolly, J.-F., Espinasse, T., Fougères, A.-L. and Ribatet, M.(2024).Spatio-temporal point process intensity estimation using zero-deflated subsampling applied to a lightning strikes dataset in France.Electronic Journal of Statistics18 5370–5404

    2024

  15. [23]

    and Lavancier, F

    Coeurjolly, J.-F. and Lavancier, F. (2019). Understanding spatial point patterns through intensity and conditional intensities.Stochastic Geometry: Modern Research Frontiers45– 85

    2019

  16. [24]

    and Amblard, P.-O

    Coeurjolly, J.-F., Mazoyer, A. and Amblard, P.-O. (2021). Monte Carlo integration of non-differentiable functions on[0, 1]ι,ι = 1,...,d , using a single determinantal point pattern defined on [0, 1]d. Electronic Journal of Statistics15 6228–6280

    2021

  17. [25]

    Coste, S. (2021). Order, fluctuations, rigidities.preprint

    2021

  18. [26]

    and Hennig, P

    Da Costa, N., Pförtner, M., Da Costa, L. and Hennig, P. (2023). Sample path regularity of Gaussian processes from the covariance kernel.arXiv preprint arXiv:2312.14886

    2023

  19. [27]

    Daley, D. J. and Vere-Jones, D. (2006). An introduction to the theory of point processes: volume I: elementary theory and methods. Springer Science & Business Media

    2006

  20. [28]

    Daley, D. J. and Vere-Jones, D. (2007). An introduction to the theory of point processes: volume II: general theory and structure. Springer Science & Business Media

    2007

  21. [29]

    Dereudre, D. (2019). Introduction to the theory of Gibbs point processes.Stochastic Geometry: Modern Research Frontiers181–229

    2019

  22. [30]

    and Leblé, T

    Dereudre, D., Flimmel, D., Huesmann, M. and Leblé, T. (2024). (Non)-hyperuniformity of perturbed lattices.arXiv preprint arXiv:2405.19881

    2024 arXiv

  23. [31]

    and Fournier, J

    Estrade, A. and Fournier, J. (2016). Number of critical points of a Gaussian random field: condition for a finite variance.Statistics & Probability Letters118 94–99

    2016

  24. [32]

    and León, J

    Estrade, A. and León, J. R. (2016). A central limit theorem for the Euler characteristic of a Gaussian excursion set.The Annals of Probability44 3849-3878

    2016

  25. [33]

    and Stecconi, M

    Gass, L. and Stecconi, M. (2024). The number of critical points of a Gaussian field: finiteness of moments.Probability Theory and Related Fields1–31

    2024

  26. [34]

    and W alster, G

    Hansen, E. and W alster, G. W. (2003). Global optimization using interval analysis: revised and expanded264. CRC press

    2003

  27. [35]

    and Lachièze-Rey, R

    Hawat, D., Gautier, G., Bardenet, R. and Lachièze-Rey, R. (2023). On estimating the structure factor of a point process, with applications to hyperuniformity.Statistics and Computing 33 61

    2023

  28. [36]

    , Penttinen, A

    Illian, J. , Penttinen, A. , Stoyan, H. and Stoyan, D. (2008). Statistical analysis and modelling of spatial point patterns. John Wiley & Sons

    2008

  29. [37]

    Ismail, M. (2005). Classical and quantum orthogonal polynomials in one variable13. Cambridge university press

    2005

  30. [38]

    Jow, D. L. , Contreras, D. , Scott, D. and Bunn, E. F. (2019). Taller in the saddle: constraining CMB physics using saddle points. Journal of Cosmology and Astroparticle Physics 2019 031

    2019

  31. [39]

    Kac, M. (1943). On the average number of real roots of a random algebraic equation.Bulletin of the American Mathematical Society49 314–320

    1943

  32. [40]

    Kratz, M. F. and León, J. R. (2006). On the second moment of the number of crossings by a stationary Gaussian process.The Annals of Probability34 1601-1607

    2006

  33. [41]

    and Lachièze-Rey, R

    Ladgham, S. and Lachièze-Rey, R. (2023). Local repulsion of planar Gaussian critical points. Stochastic Processes and their Applications166 104221

    2023

  34. [42]

    Lantuéjoul, C. (2013). Geostatistical simulation: models and algorithms. Springer Science & Business Media

    2013

  35. [43]

    Larson, D. L. and W andelt, B. D.(2004). The hot and cold spots in the Wilkinson microwave anisotropy probe data are not hot and cold enough.The Astrophysical Journal613 L85

    2004

  36. [44]

    Larsson-Cohn, L. (2002). Lp-norms of Hermite polynomials and an extremal problem on Wiener chaos.Arkiv för Matematik40 133–144

    2002

  37. [45]

    and Rubak, E

    Lavancier, F., Møller, J. and Rubak, E. (2015). Determinantal point process models and statisticalinference. Journal of the Royal Statistical Society Series B: Statistical Methodology 77 853–877

    2015

  38. [46]

    and Talagrand, M

    Ledoux, M. and Talagrand, M. (2013). Probability in Banach Spaces: isoperimetry and processes. Springer Science & Business Media. 56

    2013

  39. [47]

    and Rychlik, I

    Lindgren, G. and Rychlik, I. (1982). Wave characteristic distributions for Gaussian waves—wave-length, amplitude and steepness.Ocean Engineering9 411–432

    1982

  40. [48]

    and McBratney, A

    Minasny, B. and McBratney, A. B. (2005). The Matérn function as a general model for soil variograms. Geoderma 128 192–207

    2005

  41. [49]

    , Syversveen, A

    Møller, J. , Syversveen, A. R. and W aagepetersen, R. P. (1998). Log Gaussian Cox processes. Scandinavian Journal of Statistics25 451–482

    1998

  42. [50]

    and W aagepetersen, R

    Møller, J. and W aagepetersen, R. P. (2003). Statistical inference and simulation for spatial point processes. Chapman and Hall/CRC

    2003

  43. [51]

    Nicolaescu, L. I. (2017). A CLT concerning critical points of random functions on a Euclidean space. Stochastic Processes and their Applications127 3412–3446

    2017

  44. [52]

    and Taqqu, M

    Peccati, G. and Taqqu, M. S. (2011). Wiener Chaos: Moments, Cumulants and Diagrams: A survey with computer implementation1. Springer Science & Business Media

    2011

  45. [53]

    and Olhede, S

    Rajala, T., Murrell, D. and Olhede, S. (2018). Detecting multivariate interactions in spatial point patterns with Gibbs models and variable selection.Journal of the Royal Statistical Society Series C: Applied Statistics67 1237–1273

    2018

  46. [54]

    Rice, S. O. (1944). Mathematical analysis of random noise.The Bell System Technical Journal 23 282–332

    1944

  47. [55]

    and Gonçalves, P

    Riedi, R. and Gonçalves, P. (2015). A strong Tauberian theorem for characteristic functions. Applied and Computational Harmonic Analysis39 552–557

    2015

  48. [56]

    Ripley, B. D. (1979). Tests of ‘randomness’ for spatial point patterns.Journal of the Royal Statistical Society Series B: Statistical Methodology41 368–374

    1979

  49. [57]

    Rivera, A. (2021). High-dimensional Monochromatic Random Waves approximate the Bargmann-Fock field. Preprint

    2021

  50. [58]

    Rudin, W. (2017). Fourier analysis on groups. Courier Dover Publications

    2017

  51. [59]

    , Giordano, M

    Sanders, D. , Giordano, M. , Benet, L. , Ferranti, L. and Hénot, O. (2024). IntervalRootFinding.jl: Library for finding the roots of a function using interval arithmetic. https://juliaintervals.github.io/dev/

    2024

  52. [60]

    , Malinowski, A

    Schlather, M. , Malinowski, A. , Menck, P. J. , Oesting, M. and Strokorb, K. (2015). Analysis, Simulation and Prediction of Multivariate Random Fields with Package RandomFields. Journal of Statistical Software63 1–25

    2015

  53. [61]

    Soulier, P. (2001). Moment bounds and central limit theorem for functions of Gaussian vectors. Statistics & Probability Letters54 193–203

    2001

  54. [62]

    Stein, M. L. (2012). Interpolation of spatial data: some theory for kriging. Springer Science & Business Media

    2012

  55. [63]

    and Gómez-Antonio, M

    Sweeney, S. and Gómez-Antonio, M. (2016). Localization and industry clustering econometrics:AnassessmentofGibbsmodelsforspatialpointprocesses. Journal of Regional Science 56 257–287

    2016

  56. [64]

    Tao, T.(2023).Analysis II: Fourth Edition.Texts and Readings in Mathematics.SpringerNature Singapore

    2023

  57. [65]

    Taylor, J. E. and Worsley, K. J. (2007). Detecting sparse signals in random fields, with an application to brain mapping.Journal of the American Statistical Association102 913–928

    2007

  58. [66]

    and Guan, Y

    W aagepetersen, R. and Guan, Y. (2009). Two-step estimation for inhomogeneous spatial point processes. Journal of the Royal Statistical Society Series B: Statistical Methodology 71 685–702

    2009

  59. [67]

    Wood, A. T. and Chan, G. (1994). Simulation of stationary Gaussian processes in [0, 1]d. Journal of Computational and Graphical Statistics3 409–432

    1994

  60. [68]

    Worsley, K. J. , Marrett, S., Neelin, P. and Evans, A. (1996). Searching scale space for activation in PET images.Human brain mapping4 74–90

    1996

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.