pith. sign in

arxiv: 2601.10420 · v1 · pith:BEXWZ2XCnew · submitted 2026-01-15 · ✦ hep-ph

On the reconstruction of kinematic distributions computed with Monte Carlo methods using orthogonal basis functions

Kirill Melnikov , Ivan Novikov , Ivan Pedron This is my paper

Pith reviewed 2026-05-21 15:03 UTC · model grok-4.3

classification ✦ hep-ph
keywords Monte Carlo integrationkinematic distributionsorthogonal basis functionsperturbative calculationslocal subtractionshistogram fluctuationsHiggs production
0
0 comments X

The pith

Kinematic distributions from Monte Carlo calculations can be rebuilt as smooth sums of orthogonal basis functions rather than binned histograms.

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

The paper shows how to reconstruct one-dimensional kinematic distributions by expanding them as a weighted sum of orthogonal basis functions, with the coefficients obtained directly from the Monte Carlo integration. This produces continuous approximations instead of discrete histograms. In high-order perturbative calculations that employ local subtractions, the approach removes the bin-to-bin statistical fluctuations that degrade conventional histograms. The authors further demonstrate that a known high-quality approximation, such as the leading-order result, can be used to construct an optimized orthonormal basis. They validate the method on toy models and on a realistic example of Higgs boson production in weak boson fusion.

Core claim

The target kinematic distribution is approximated by a finite sum of orthogonal basis functions whose coefficients are computed as Monte Carlo averages of the integrand multiplied by each basis function. This yields a smooth reconstruction that eliminates the statistical fluctuations between histogram bins, which are especially pronounced in perturbative calculations with local subtractions.

What carries the argument

The central mechanism is the orthonormal expansion of the distribution, where each coefficient is evaluated by Monte Carlo integration over the high-dimensional phase space and the basis is optionally optimized using a leading-order approximation to the same distribution.

If this is right

  • Perturbative predictions at next-to-leading order and beyond become less sensitive to Monte Carlo statistical errors when plotted as continuous functions.
  • The reconstructed distributions can be differentiated or integrated analytically for use in further calculations without re-binning.
  • Optimized bases built from leading-order results accelerate convergence of the expansion for higher-order corrections.
  • The method applies directly to processes with local subtraction terms, such as Higgs production in weak boson fusion.

Where Pith is reading between the lines

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

  • The same expansion idea could be extended to selected multi-dimensional distributions if suitable tensor-product bases are available.
  • Combining the basis reconstruction with adaptive Monte Carlo sampling might further reduce the number of events needed for a given precision.
  • The approach could be tested on distributions that contain sharp resonances or thresholds to quantify any truncation effects near those features.

Load-bearing premise

A finite truncation of the orthogonal basis expansion can represent the target kinematic distribution with sufficient accuracy and without introducing significant bias or truncation artifacts that affect the final physics results.

What would settle it

Generate a high-statistics reference histogram for the same distribution and check whether the basis-function reconstruction deviates systematically from it in any kinematic region or introduces oscillations absent from the reference.

Figures

Figures reproduced from arXiv: 2601.10420 by Ivan Novikov, Ivan Pedron, Kirill Melnikov.

Figure 1
Figure 1. Figure 1: Legendre moments ⟨f, ek⟩ of the normal distribu￾tion f = N (0, 1) (see Eq. (22) for the definition) on the in￾terval [−1, 2]. The blue “Accurate” crosses are the true Leg￾endre moments, calculated with high precision. The black data points are the moments calculated in a toy Monte-Carlo with n = 105 samples. The orange dashed line is the estimate ⟨f, ek⟩ ∼ crk of the typical magnitude of the Legendre mo￾me… view at source ↗
Figure 2
Figure 2. Figure 2: Normal distribution (2π) −1/2 exp(−x 2 /2) reconstructed from n samples using a histogram or Legendre moments. All plots, except for the bottom-left one, compare the two approximations on the interval [−1, 2]. The bottom-left plot shows the approximation on the interval [−1.5, 5]. The plots in the top row differ in the number of Monte-Carlo samples n. The histograms in the right column differ in the bin wi… view at source ↗
Figure 3
Figure 3. Figure 3: Examples of distributions that are more difficult to approximate using Legendre moments. The distribution on [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Distribution of transverse momentum p⊥,H of Higgs boson produced in WBF, reconstructed using histograms (left) or Legendre moments (right). The blue, orange, and black lines correspond to the leading-, next-to-leading, and next-to-next￾to-leading order result. The bottom panels show relative deviations from the leading-order one. For Legendre moments the number in brackets indicates the number of moments r… view at source ↗
Figure 5
Figure 5. Figure 5: Distribution of rapidity yH of Higgs boson produced in WBF, reconstructed using histograms (left) or Legendre moments (right). The blue, orange, and black lines correspond to the leading-, next-to-leading, and next-to-next-to-leading order result. The bottom panels show relative deviations from the leading-order one. For Legendre moments the number in brackets indicates the number of moments retained after… view at source ↗
Figure 6
Figure 6. Figure 6: Examples of bin-to-bin fluctuations in some infrared-subtracted double-real NNLO corrections to WBF. The solid [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Basis functions constructed from an initial approxi [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Distribution of transverse momentum of the Higgs boson approximated using using Legendre moments (left) and [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The left plots shows the distribution of transverse momentum of the Higgs boson in weak-boson fusion approximated [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗
read the original abstract

Reconstruction of one-dimensional kinematic distributions from calculations based on high-dimensional Monte-Carlo integration is a standard problem in high-energy physics. Traditionally, this is done by collecting randomly-generated events in histograms. In this article, we explore an alternative approach, whose main idea is to approximate the target distribution by a weighted sum of orthogonal basis functions whose coefficients are calculated using the Monte-Carlo integration. This method has the advantage of directly yielding smooth approximations to target distributions. Furthermore, in the context of high-order perturbative calculations with local subtractions, it eliminates the so-called bin-to-bin fluctuations, which often severely affect the quality of conventional histograms. We also demonstrate that the availability of a high-quality approximation to the target distribution, for example the leading-order result in the perturbative expansion, can be exploited to construct an optimized orthonormal basis. We compare the performance of this method to conventional histograms in both toy-model and real Monte-Carlo settings, applying it to Higgs boson production in weak boson fusion as an example.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper proposes reconstructing one-dimensional kinematic distributions from high-dimensional Monte Carlo integrations by expanding the target in a sum of orthogonal basis functions, with coefficients obtained directly from the MC integrator. This yields smooth approximations and, in the presence of local subtractions at higher perturbative orders, removes bin-to-bin fluctuations that plague conventional histograms. The authors further show how a high-quality approximation (e.g., the leading-order result) can be used to construct an optimized orthonormal basis and compare the method against histograms in toy models and in a realistic Higgs weak-boson-fusion calculation.

Significance. If the truncation error can be shown to remain below the statistical and perturbative precision of the underlying calculation, the technique would supply a practical tool for producing smooth, fluctuation-free differential distributions in NLO and NNLO Monte Carlo programs that employ local subtraction schemes. The availability of an optimized basis derived from a lower-order result is a concrete strength that could be adopted more broadly.

major comments (3)
  1. [§4] §4 (toy-model section): the comparisons are presented without quantitative tables or plots that report the integrated absolute deviation or Kolmogorov-Smirnov distance between the truncated basis reconstruction and the known analytic target, making it impossible to verify that truncation artifacts remain smaller than the Monte Carlo statistical uncertainty.
  2. [§5] §5 (WBF Higgs example): the claim that bin-to-bin fluctuations are eliminated must be accompanied by a direct comparison, for at least one distribution, of the basis-reconstructed result against a high-statistics reference histogram or against the unsubtracted integrand; without such a benchmark it is unclear whether the global smoothing introduced by a finite basis order distorts the large positive/negative cancellations that characterize local subtraction terms.
  3. [§3.2] §3.2 (basis optimization): the procedure for choosing the truncation order N is not specified; the manuscript should state an a-priori criterion (e.g., monitoring the size of the last few coefficients or the change in the reconstructed distribution when N is increased by one) that guarantees the residual bias lies below the target precision of the physics analysis.
minor comments (2)
  1. [§2] The notation for the orthogonal polynomials (or other basis functions) and the inner-product measure should be introduced once in §2 and used consistently thereafter.
  2. Figure captions should explicitly state the truncation order N employed for each curve and whether error bands include only Monte Carlo statistics or also an estimate of truncation uncertainty.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the careful reading and the constructive suggestions. We address each major comment below and will revise the manuscript to incorporate the requested quantitative validations, benchmarks, and explicit selection criteria.

read point-by-point responses

  1. Referee: [§4] §4 (toy-model section): the comparisons are presented without quantitative tables or plots that report the integrated absolute deviation or Kolmogorov-Smirnov distance between the truncated basis reconstruction and the known analytic target, making it impossible to verify that truncation artifacts remain smaller than the Monte Carlo statistical uncertainty.

    Authors: We agree that quantitative error metrics would make the validation more rigorous. In the revised manuscript we will add a table in §4 that reports both the integrated absolute deviation and the Kolmogorov-Smirnov distance between the truncated basis expansion and the known analytic target for several values of N. These numbers will be compared directly to the Monte Carlo statistical uncertainties quoted in the same section. revision: yes

  2. Referee: [§5] §5 (WBF Higgs example): the claim that bin-to-bin fluctuations are eliminated must be accompanied by a direct comparison, for at least one distribution, of the basis-reconstructed result against a high-statistics reference histogram or against the unsubtracted integrand; without such a benchmark it is unclear whether the global smoothing introduced by a finite basis order distorts the large positive/negative cancellations that characterize local subtraction terms.

    Authors: We will include the requested benchmark in the revised §5. For the leading differential distribution we will add a direct overlay of the basis-reconstructed result against both (i) a high-statistics reference histogram generated with an independent, much larger Monte Carlo sample and (ii) the unsubtracted integrand. The comparison will be shown at the same binning used in the original figure so that any residual distortion from the finite-order smoothing can be quantified. revision: yes

  3. Referee: [§3.2] §3.2 (basis optimization): the procedure for choosing the truncation order N is not specified; the manuscript should state an a-priori criterion (e.g., monitoring the size of the last few coefficients or the change in the reconstructed distribution when N is increased by one) that guarantees the residual bias lies below the target precision of the physics analysis.

    Authors: We will expand §3.2 to state an explicit a-priori criterion. The truncation order N is chosen such that the absolute value of the highest retained coefficient satisfies |c_N| < ε, where ε is set to 10 % of the target statistical precision of the distribution. In addition, we will verify that increasing N by one changes the reconstructed distribution by less than the same tolerance. Both the coefficient threshold and the stability check will be documented with a short algorithmic description and illustrated for the toy-model case. revision: yes

Circularity Check

0 steps flagged

No circularity: coefficients computed directly via Monte Carlo integration on independent basis

full rationale

The paper's core method computes expansion coefficients for an orthogonal basis directly from Monte Carlo integration of the target integrand. This is a standard, non-self-referential numerical procedure equivalent to projecting the distribution onto the basis. The use of a leading-order approximation to optimize the basis is an external input from perturbation theory, not derived from the method itself. Comparisons to histograms in toy models and WBF examples are direct numerical tests without fitted parameters renamed as predictions or load-bearing self-citations. The derivation chain remains independent of its outputs.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The approach rests on the mathematical completeness of the chosen orthogonal basis for the target distribution and on the statistical convergence of Monte Carlo estimates for the expansion coefficients; both are standard but must be verified for each application.

free parameters (1)
  • truncation order of the basis expansion
    The number of retained basis functions is a practical cutoff that must be chosen to balance accuracy and computational cost.
axioms (1)
  • domain assumption The target distribution lies in the span of the selected orthogonal basis to within acceptable truncation error
    Invoked when the method is asserted to produce accurate smooth approximations.

pith-pipeline@v0.9.0 · 5704 in / 1236 out tokens · 123247 ms · 2026-05-21T15:03:49.781167+00:00 · methodology

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

23 extracted references · 23 canonical work pages · 10 internal anchors

  1. [1]

    James, Monte Carlo Theory and Practice, Rept

    F. James, Monte Carlo Theory and Practice, Rept. Prog. Phys.43, 1145 (1980)

    work page 1980

  2. [2]

    G. P. Lepage, A New Algorithm for Adaptive Multidi- mensional Integration, J. Comput. Phys.27, 192 (1978)

    work page 1978

  3. [3]

    Navaset al.(Particle Data Group), Review of particle physics, Phys

    S. Navaset al.(Particle Data Group), Review of particle physics, Phys. Rev. D110, 030001 (2024)

    work page 2024

  4. [4]

    General-purpose event generators for LHC physics

    A. Buckleyet al., General-purpose event genera- tors for LHC physics, Phys. Rept.504, 145 (2011), arXiv:1101.2599 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  5. [5]

    B. W. Silverman,Density estimation for statistics and data analysis(Chapman and Hall, London, 1986)

    work page 1986

  6. [6]

    D. G. Truhlar and N. C. Blais, Legendre moment method for calculating differential scattering cross sections from classical trajectories with monte carlo initial conditions, The Journal of Chemical Physics 67, 1532 (1977), https://pubs.aip.org/aip/jcp/article- pdf/67/4/1532/18907999/1532 1 online.pdf

    work page 1977

  7. [7]

    Three-jet cross sections to next-to-leading order

    S. Frixione, Z. Kunszt, and A. Signer, Three jet cross- sections to next-to-leading order, Nucl. Phys. B467, 399 (1996), arXiv:hep-ph/9512328

    work page internal anchor Pith review Pith/arXiv arXiv 1996

  8. [8]

    A General Algorithm for Calculating Jet Cross Sections in NLO QCD

    S. Catani and M. H. Seymour, A General algorithm for calculating jet cross-sections in NLO QCD, Nucl. Phys. B485, 291 (1997), [Erratum: Nucl.Phys.B 510, 503–504 (1998)], arXiv:hep-ph/9605323

    work page internal anchor Pith review Pith/arXiv arXiv 1997

  9. [9]

    Antenna Subtraction at NNLO

    A. Gehrmann-De Ridder, T. Gehrmann, and E. W. N. Glover, Antenna subtraction at NNLO, JHEP09, 056, arXiv:hep-ph/0505111

    work page internal anchor Pith review Pith/arXiv arXiv

  10. [10]

    A novel subtraction scheme for double-real radiation at NNLO

    M. Czakon, A novel subtraction scheme for double-real radiation at NNLO, Phys. Lett. B693, 259 (2010), arXiv:1005.0274 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  11. [11]

    Fully differential VBF Higgs production at NNLO

    M. Cacciari, F. A. Dreyer, A. Karlberg, G. P. Salam, and G. Zanderighi, Fully Differential Vector-Boson- Fusion Higgs Production at Next-to-Next-to-Leading Or- der, Phys. Rev. Lett.115, 082002 (2015), [Erratum: Phys.Rev.Lett. 120, 139901 (2018)], arXiv:1506.02660 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  12. [12]

    Jet production in the CoLoRFulNNLO method: event shapes in electron-positron collisions

    V. Del Duca, C. Duhr, A. Kardos, G. Somogyi, Z. Sz˝ or, Z. Tr´ ocs´ anyi, and Z. Tulip´ ant, Jet production in the CoLoRFulNNLO method: event shapes in electron- positron collisions, Phys. Rev. D94, 074019 (2016), arXiv:1606.03453 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  13. [13]

    Nested soft-collinear subtractions in NNLO QCD computations

    F. Caola, K. Melnikov, and R. R¨ ontsch, Nested soft- collinear subtractions in NNLO QCD computations, Eur. Phys. J. C77, 248 (2017), arXiv:1702.01352 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  14. [14]

    Local Analytic Sector Subtraction at NNLO

    L. Magnea, E. Maina, G. Pelliccioli, C. Signorile- Signorile, P. Torrielli, and S. Uccirati, Local analytic sector subtraction at NNLO, JHEP12, 107, [Erratum: JHEP 06, 013 (2019)], arXiv:1806.09570 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  15. [15]

    V. D. Duca, C. Duhr, L. Fek´ esh´ azy, F. Guadagni, P. Mukherjee, G. Somogyi, F. Tramontano, and S. V. Thurenhout, Nnlocal: Completely local subtractions (2026), arXiv:2601.02033 [hep-ph]

    work page arXiv 2026

  16. [16]

    Szeg¨ o,Orthogonal Polynomials(American Mathemat- ical Society, New York City, 1939)

    G. Szeg¨ o,Orthogonal Polynomials(American Mathemat- ical Society, New York City, 1939)

    work page 1939

  17. [17]

    J. P. Boyd,Chebyshev and Fourier Spectral Methods, 2nd ed. (Dover, Mineola, New York, 2001)

    work page 2001

  18. [18]

    Treating the b quark distribution function with reliable uncertainties

    Z. Ligeti, I. W. Stewart, and F. J. Tackmann, Treating the b quark distribution function with reliable uncertain- ties, Phys. Rev. D78, 114014 (2008), arXiv:0807.1926 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2008

  19. [19]

    C. W. Clenshaw, A note on the summation of chebyshev series, Mathematics of Computation9, 118 (1955)

    work page 1955

  20. [20]

    Wang and S

    H. Wang and S. Xiang, On the convergence rates of leg- endre approximation, Mathematics of computation81, 13 861 (2012)

    work page 2012

  21. [21]

    Asteriadis, F

    K. Asteriadis, F. Caola, K. Melnikov, and R. R¨ ontsch, NNLO QCD corrections to weak boson fusion Higgs bo- son production in the H→b band H→WW ∗ →4l decay channels, JHEP02, 046 (2022), arXiv:2110.02818 [hep-ph]

    work page arXiv 2022

  22. [22]

    Bierlich et al.,Robust Independent Validation of Experiment and Theory: Rivet version 3, SciPost Phys.8 (2020) 026, arXiv:1912.05451 [hep-ph]

    C. Bierlichet al., Robust Independent Validation of Ex- periment and Theory: Rivet version 3, SciPost Phys.8, 026 (2020), arXiv:1912.05451 [hep-ph]

    work page arXiv 2020

  23. [23]

    Bothmannet al.(Sherpa), Event generation with Sherpa 3, JHEP12, 156, arXiv:2410.22148 [hep-ph]

    E. Bothmannet al.(Sherpa), Event generation with Sherpa 3, JHEP12, 156, arXiv:2410.22148 [hep-ph]

    work page arXiv