Pith. sign in

REVIEW 2 major objections 4 minor 49 references

2D Nondirect Product Discrete Variable Representation for Schr\"odinger Equation with Nonseparable Angular Variables

T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read A wave-function basis built on Popov or Lebedev unit-sphere cubatures converges substantially faster than the old Gaussian-product angular grid for Schrödinger equations with nonseparable angular variables.

desk verdict A solid, incremental extension of npDVR to Lebedev and Popov cubatures with a convincing single-system benchmark, but the claimed correlation between cubature efficiency and convergence is not rigorously established and the quadrature error for the high-l tail is left unquantified. read the letter →

arxiv 1909.01578 v2 pith:MUPYBVNB submitted 2019-09-04 physics.comp-ph physics.atom-ph

Sara Shadmehri , Shahpoor Saeidian , Vladimir S. Melezhik This is my paper · ORCID
classification physics.comp-phphysics.atom-ph
keywords nondirectproductDVRLebedevquadraturePopovunit-spherecubatureangulardiscretizationhydrogenatomcrossedfieldsefficiencycoefficientSchrödingerequation
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

This paper develops a nondirect product discrete variable representation (npDVR) for quantum problems whose angular variables cannot be separated, replacing the old direct product of 1D Gaussian quadratures with Lebedev or Popov 2D cubatures on the unit sphere. It constructs orthogonalized basis functions on the cubature grid points and tests them on the hydrogen atom in arbitrarily oriented electric and magnetic fields, a genuinely nonseparable three-dimensional problem. The paper finds that both cubature-based schemes converge substantially faster than the Gaussian-product scheme at the same number of grid points, and that the Popov-based npDVR converges fastest, consistent with its higher efficiency coefficient. If correct, this gives quantum dynamics a cheaper angular discretization that also avoids the pole-clustering of product grids, making time-dependent few-electron simulations more feasible.

What carries the argument

The central object is the cubature-orthogonalized basis Φν(Ω)=∑μ SνμYμ(Ω), obtained by diagonalizing the overlap matrix built from spherical harmonics on the Lebedev or Popov grid points. The key identity is the orthogonality and completeness pair (12)-(13) on the grid, which makes the potential matrix diagonal and yields the L2 matrix formula (15). The paper ties the method's performance to the cubature efficiency coefficient η, the ratio of exactly integrated spherical harmonics to the number of grid degrees of freedom, and argues that larger η should give a better approximation for a fixed number of basis functions.

What would settle it

Take a fixed N-point Popov grid (e.g. N=12), build the overlap matrix (10) up to the stated ltop, and compare its eigenvalues with the ideal values 1 and 0; any nonzero eigenvalue beyond N, or significant deviation of the N kept eigenvalues from unity, would show the grid does not support the claimed completeness and should worsen convergence. The same comparison using a high-order reference quadrature for the exact scalar products would settle it.

Watch

Extended reading notes

Core claim

The central claim is that a 2D npDVR built from invariant cubatures on the unit sphere outperforms the earlier npDVR built from a direct product of 1D Gaussian quadratures. The paper constructs the basis by extending the spherical-harmonic set until the overlap matrix (10), evaluated with the cubature weights, has exactly N nonzero eigenvalues; the normalized eigenvectors define functions Φν that are orthogonal and complete on the grid. With these functions the angular kinetic-energy operator L2 takes the explicit form (15) and the potential remains diagonal. Tested on the hydrogen atom in crossed magnetic and electric fields, the paper reports clearly faster convergence for both the Lebedev and Popov grids than for the Gaussian-product npDVR, with the Popov grid giving accurate results already at N=12. The paper concludes that convergence speed tracks the cubature efficiency coefficient η=(n+1)2/(3N).

Load-bearing premise

The load-bearing premise is that the Lebedev or Popov cubature weights compute the spherical-harmonic overlap matrix accurately enough up to ltop so that the N kept eigenvectors really are orthogonal and complete on the grid; if the cubature errors for high-l products are not small, the whole npDVR basis is only approximately orthogonal.

Editorial extensions

If this is right

  • For the hydrogen spectrum in crossed fields, the Popov-based npDVR reaches accurate energies with far fewer angular grid points than the Gaussian-product scheme, with N=12 already giving good results.
  • The absence of pole clustering in the Lebedev and Popov grids removes a time-step restriction in time-dependent Schrödinger implementations.
  • The construction works with any invariant cubature on the unit sphere, so the efficiency-coefficient ranking provides a direct criterion for choosing the angular grid.
  • The method extends naturally to Schrödinger problems with nonseparable angular variables where partial-wave expansions become inefficient due to strong coupling.
  • The reported correlation between η and convergence gives a practical rule: for a fixed number of grid points, the cubature with the highest η should be preferred.

Reading between the lines

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

  • The efficiency-coefficient heuristic could be used as a design rule for pruned angular grids in higher-dimensional problems, where direct-product grids become prohibitively large.
  • The overlap-matrix diagonalization used here can serve as a general diagnostic: before committing to dynamics on any candidate angular grid, one can check that the number of near-unit eigenvalues matches the grid size.
  • If the reported acceleration holds more broadly, similar gains should appear in time-dependent two-electron problems, a direction the paper only hints at in its conclusion.
  • The high-l tail of the extended harmonic set carries the quadrature error; testing the method on potentials with stronger high-l coupling would show how far the ltop extension can be pushed before the approximation degrades.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper develops a nondirect product discrete variable representation (npDVR) for the angular part of the Schrödinger equation when the angular variables are nonseparable. The new basis is constructed by diagonalizing the overlap matrix of spherical harmonics computed on Lebedev or Popov cubature grids for the unit sphere, rather than on the direct product of one-dimensional Gaussian grids used previously. The authors test the method by computing bound states of a hydrogen atom in mutually tilted electric and magnetic fields, comparing the Gaussian-product, Lebedev, and Popov schemes. They report that the Popov-based npDVR converges fastest, followed by the Lebedev-based one, and they attribute this ordering to the cubature efficiency coefficient η.

Significance. If the reported acceleration is robust, the method is a genuinely useful tool for quantum dynamics with nonseparable angular degrees of freedom: it removes the polar clustering of the old product grid and can reach accurate energies with substantially fewer angular points. The paper has real strengths: the basis construction is described in enough detail to be reproduced, the hydrogen-in-crossed-fields test is a standard benchmark with an independent earlier value from Ref. [2], and the tables give the numerical data in full rather than as anonymous convergence plots. The main limitation is that the central correlation claim is supported by a single example and by heuristic reasoning, while the underlying quadrature-exactness issue for the high-l part of the basis is not analyzed.

major comments (2)
  1. [Sec. 2, Eqs. (10)-(15)] The overlap matrix in Eq. (10) is evaluated with cubature weights that are exact only for products of spherical harmonics with angular momenta up to lmax, whereas the basis includes harmonics up to ltop > lmax (Table 1 gives, e.g., N=110 with lmax=8 and ltop=12, and N=104 with lmax=8 and ltop=11). The quadrature error in the Gram matrix and in the L2 matrix (15) is never quantified, and the condition number of the overlap matrix is not reported. Since ltop/lmax is not shown to tend to unity as N grows, the high-l component of the basis does not automatically become negligible; consequently it is not established that the quadrature error decays with N. This gap is load-bearing because a poor discrete representation of the high-l tail could cause the convergence curves in Fig. 1 to saturate or could change the relative ranking of the three schemes at stricter tolerances. The revision should either bound the quadrature error, report the spectrum and conditioning of the overlap matrix, or demonstrate numerically that the error decays with N.
  2. [Sec. 3, Tables 2 and 3] The claimed direct correlation between η and convergence is not fully supported by the tabulated data because the convergence is non-monotonic in N. In Table 2 the Gaussian scheme gives -0.1307377 at N=49 but -0.1307454 at N=50 and N=70, and in Table 3 it gives -0.1304448 at N=30 and N=36 but -0.1302756 at N=40, followed by -0.1302755 at N=49 and -0.1302756 at N=50 and N=70. These jumps mean that comparisons at a fixed N can depend on the chosen tolerance and on which side of a jump the comparison is made. The paper should define a convergence criterion (e.g., deviation from a converged reference energy) and show that the ordering 'Gaussian slower than Lebedev slower than Popov' holds under that criterion, ideally for more than one state and more than one field configuration.
minor comments (4)
  1. [Fig. 1] The thin horizontal lines labeled 'most accurate value' are apparently obtained from the large-N limit of the same schemes, but the caption does not state the reference value used or the tolerance for convergence; please make this explicit.
  2. [Tables 2 and 3] The arrangement of the Gaussian columns for Nφ=3,5,7 is confusing, and the N=49/N=50 entries should be explained in terms of the underlying Nθ, Nφ grid parameters, especially because those two rows show the largest non-monotonic jumps.
  3. [Sec. 2, text near Eq. (9)] The statement that the Lebedev or Popov cubature can exactly integrate 'the square of spherical harmonics up to a certain angular momentum (lmax)' is imprecise: the cubatures are exact for all products of pairs of spherical harmonics whose angular momenta do not exceed lmax, which is the property actually needed in Eq. (10).
  4. [References] Several reference-list typos should be corrected: 'Phyladelphia' and 'Schience' in Ref. [45], 'editted' in Ref. [4], and 'Retreived' in Ref. [49].

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the npDVR construction is self-contained and the convergence claims are measured against independent computed energies and an earlier benchmark.

full rationale

The paper's derivation chain builds the npDVR basis from the Lebedev or Popov cubature weights via Eqs. (10)-(15): an overlap matrix is formed from spherical harmonics on the cubature grid, diagonalized, and the resulting orthonormal combinations define the DVR basis and the L2 operator. No parameter in this construction is fitted to the energies that are later reported. The central claim, that Popov and Lebedev cubatures accelerate convergence relative to product Gaussian grids, is supported by independent eigenvalue calculations of the hydrogen atom in crossed fields; the efficiency coefficient eta is an independent property of each cubature (Table 1), not a fit to the computed spectra. The weak-field comparison uses the earlier calculation of Melezhik [2] as an external benchmark, but that prior result is not an input to the new basis construction and the new method is also assessed by its own convergence to the large-N limit. The unresolved question about quadrature error for l > lmax is a numerical accuracy concern, not a circularity: it does not make any predicted quantity equal to an input by construction. The correlation between eta and observed convergence is an empirical finding, not an equation identity. No load-bearing argument reduces to a self-citation, and no fitted input is renamed as a prediction. Therefore the appropriate circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard numerical assumptions about DVR convergence and on the exactness properties of the cited cubatures. No parameters are fitted to the target energies, and no new physical entities are introduced.

assumptions (4)
  • domain assumption Lebedev and Popov cubatures exactly integrate all spherical harmonics up to order n and all products of pairs with angular momentum up to lmax (Table 1).
    This exactness is the basis for the overlap matrix in Eq. (10); it is taken from the cited literature on cubatures [19-24] and not re-derived here.
  • domain assumption The npDVR expansion (2) converges to the solution of the Schrödinger equation as N increases.
    The paper assumes the basis set is complete in the limit; this is standard for DVR methods.
  • domain assumption The radial discretization error is negligible compared to the angular npDVR error.
    Stated in Section 3: the radial step and boundary rm are fixed so the integration error does not exceed the angular approximation error.
  • standard math The efficiency coefficient eta = (n+1)^2/(3N) is a meaningful measure of cubature quality for DVR convergence.
    Used to rank grids in Table 1 and to interpret convergence; the formula is standard (McLaren [43]).

how reviews work

0 comments
Cite this review

Pith. "Pith review of 2D Nondirect Product Discrete Variable Representation for Schr\"odinger Equation with Nonseparable Angular Variables." pith.science (2026). https://pith.science/paper/MUPYBVNB

@misc{pith2026190901578,
  author       = {Pith},
  title        = {Pith review of: 2D Nondirect Product Discrete Variable Representation for Schr\"odinger Equation with Nonseparable Angular Variables},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MUPYBVNB}},
  note         = {Machine review of arXiv:1909.01578}
}
read the original abstract

We develop a nondirect product discrete variable representation (npDVR) for treating quantum dynamical problems which involve nonseparable angular variables. The npDVR basis is constructed on spherical functions orthogonalized on the grids of the Lebedev or Popov 2D quadratures for the unit sphere instead of the direct product of 1D quadrature rules. We compare our computational scheme with the old one that used the product of 1D Gaussian quadratures in terms of their convergence and efficiency by calculating, as an example, the spectrum of a hydrogen atom in the magnetic and electric fields arbitrarily oriented to one another. The use of the npDVR based on the Lebedev or Popov 2D quadratures substantially accelerates the convergence of the computational scheme. Moreover, we get the fastest convergence with the npDVR based on the Popov quadratures, which has the largest efficiency coefficient.

Figures

Figures reproduced from arXiv: 1909.01578 by the authors.

Figure 1
Figure 1. The ground-state energies of a hydrogen atom in the external magnetic β = 2.0 and electric γ = 0.5, 1.0, 1.5, 2.0 fields perpendicular to one another (α = π/2). The energies are calculated using three computational schemes with different npDVR bases (see text). The thin horizontal black lines show the most accurate value obtained for each case. the case of the ground state, the performed analysis demonstrates a dire… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 49 canonical work pages

  1. [2]

    Melezhik V S 1993 Phys. Rev. A 48 4528

    work page 1993

  2. [1]

    Melezhik V S 1991 J. Comput. Phys. 92 67

    work page 1991

  3. [3]

    Melezhik V S 1997 Phys. Lett. A 230 203

    work page 1997

  4. [4]

    Melezhik V S 1998 A Computational Method for Quantum Dynamics of Three-Dimen sional Atom in Strong Fields in Atoms and Molecules in Strong External Fields editted by Schmelcher P and Schweizer W (New York: Plenum) p 89

    work page 1998

  5. [5]

    Melezhik V S and Baye D 1999 Phys. Rev. C 59 3232

    work page 1999

  6. [6]

    Wang X -G and Carrington T Jr 2008 J. Chem. Phys. 128 194109

    work page 2008

  7. [7]

    Melezhik V S and Schmelcher P 2000 Phys. Rev. Lett. 84 1870

    work page 2000

  8. [8]

    Melezhik V S and Baye D 2001 Phys. Rev. C 64 054612

    work page 2001

Show all 49 references
  1. [9]

    Capel P, Baye D and Melezhik V S 2003 Phys. Rev. C 68 014612

    2003

  2. [10]

    Melezhik V S and Hu C Yu 2003 Phys. Rev. Lett. 90 083202

    2003

  3. [11]

    Kim J I, Melezhik V S and Schmelcher P 2006 Phys. Rev. Lett. 97 193203

    2006

  4. [12]

    Saeidian S, Melezhik V S and Schmelcher P 2008 Phys. Rev. A 77 042721

    2008

  5. [13]

    Melezhik V S and Schmelcher P 2009 New J. Phys. 11 073031

    2009

  6. [14]

    Melezhik V S and Schmelcher P 2011 Phys. Rev. A 84 042712

    2011

  7. [15]

    Giannakeas P, Melezhik V S and Schmelcher P 2011 Phys. Rev. A 84 023618

    2011

  8. [16]

    Shadmehri S, Saeidian S and Melezhik V S 2016 Phys. Rev. A 93 063616

    2016

  9. [17]

    Melezhik V S 2014 Phys. Atom. Nucl. 77 446

    2014

  10. [18]

    Melezhik V S, Kim J I and Schmelcher P 2007 Phys. Rev. A 76 053611

    2007

  11. [19]

    Lebedev V I 1975 Zh. Vychisl. Mat. Mat. Fiz. 15 48

    1975

  12. [20]

    Lebedev V I 1976 Zh. Vychisl. Mat. Mat. Fiz. 16 293

    1976

  13. [21]

    Lebedev V I 1977 Sib. Mat. Zh. 18 132

    1977

  14. [22]

    Lebedev V I and Skorokhodov A L 1992 Russ. Acad. Sci. Dokl. Math. 45 587 2D npDVR on np grids 11

    1992

  15. [23]

    Lebedev V I and Laikov D N 1999 Russ. Acad. Sci. Dokl. Math. 59 477

    1999

  16. [24]

    Popov A S 1994 Russ. J. Numer. Anal. M. 9 535

    1994

  17. [25]

    Yu H-G 2017 J. Chem. Phys. 147 094101

    2017

  18. [26]

    Smolyak S A 1963 Sov. Math. Dokl. 4 240

    1963

  19. [27]

    Avila G and Carrington T Jr 2009 J. Chem. Phys. 131 174103

    2009

  20. [28]

    Avila G and Carrington T Jr 2011 J. Chem. Phys. 134 054126

    2011

  21. [29]

    Avila G and Carrington T Jr 2011 J. Chem. Phys. 135 064101

    2011

  22. [30]

    Brown J and Carrington T Jr 2016 J. Chem. Phys. 145 144104

    2016

  23. [31]

    Acta A 119 18

    Lauvergnat D and Nauts A 2014 Spectrochim. Acta A 119 18

    2014

  24. [32]

    Larsson H R, Harke B and Tannor D J 2016 J. Chem. Phys 145 204108

    2016

  25. [33]

    Yu H-G 2005 J. Chem. Phys. 122 164107

    2005

  26. [34]

    Dickinson A S and Certain P R 1968 J. Chem. Phys. 49 4209

    1968

  27. [35]

    Lill J V, Parker G A and Light J C 1982 Chem. Phys. Lett. 89 483

    1982

  28. [36]

    Light J C and Carrington T Jr 2000 Adv. Chem. Phys. 114 263

    2000

  29. [37]

    Baye D and Heenen P-H 1986 J. Phys. A 19 2041

    1986

  30. [38]

    Status Solidi B 243 1095

    Baye D 2006 Phys. Status Solidi B 243 1095

    2006

  31. [39]

    Leforestier C 1991 J. Chem. Phys. 94 6388

    1991

  32. [40]

    Colbert D T and Miller W H 1991 J. Chem. Phys. 96 1982

    1991

  33. [41]

    Corey G C and Tromp J W 1995 J. Chem. Phys. 103 1812

    1995

  34. [42]

    Sukiasyan S and Meyer H-D 2001 J. Phys. Chem. A 105 2604

    2001

  35. [43]

    McLaren A D 1963 Math. Comput. 17, 361

    1963

  36. [44]

    Sobolev S L 1962 Soviet Math. Dokl. 3, 1307

    1962

  37. [45]

    Sobolev S L 1992 Cubature formulas and modern analysis (Phyladelphia: Gordon and Breach Schience Publishers)

    1992

  38. [46]

    Ahrens C and Beylkin G 2009 Proc. R. Soc. 465 3103

    2009

  39. [47]

    Haxton D J 2007 J. Phys. B 40 4443

    2007

  40. [48]

    108 01008

    Melezhik V S 2016 EPJ Web of Conf. 108 01008

    2016

  41. [49]

    Burkardt J 2010 Sphere Lebedev Rule, Quadrature Rules for the Unit Sphere , Retreived from https://people.sc.fsu.edu/~jburkardt/c_src/sphere_lebedev_rule/sphere_lebedev_rule.html

    2010

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.