Pith. sign in

REVIEW 3 major objections 3 minor 33 references

A Hybrid High-Order Method for the Gross--Pitaevskii Eigenvalue Problem

T0 review · 3 major / 3 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper introduces an HHO discretisation of the Gross–Pitaevskii ground state that converges at optimal order and, in a modified lowest-order variant, yields guaranteed lower energy bounds directly, without post-processing.

desk verdict Solid HHO convergence analysis for the GP-EVP, but the headline lower-bound claim overreaches: for smooth potentials the numerics freeze a coarse piecewise-constant potential, so the asymptotic exactness advertised in the abstract is not established. read the letter →

arxiv 2506.19944 v1 pith:YLDKK6QZ submitted 2025-06-24 math.NA cs.NA

Moritz Hauck , Yizhou Liang This is my paper · ORCID
classification math.NAcs.NA MSC 65N1265N1565N2565N30
keywords Gross-Pitaevskiieigenvalueproblemhybridhigh-ordermethodguaranteedlowerenergyboundsapriorierroranalysisBose-EinsteincondensatesnonlinearJensen'sinequality
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 sets out to prove that a hybrid high-order (HHO) discretisation—using cell and face polynomials linked by a reconstruction operator—approximates the ground state of the Gross–Pitaevskii eigenvalue problem at optimal order in the state, the chemical potential, and the energy. Its second goal is to construct a modified lowest-order version whose computed energy is a guaranteed lower bound on the true ground-state energy, obtained directly from the discrete solution without any post-processing. This matters because conforming finite element methods can only overestimate the ground-state energy; a provable and accurate lower bound turns an energy computation into a two-sided certificate. The paper proves L2 convergence of order $h^{r+2}$, H1 convergence of order $h^{r+1}$, energy convergence of order $h^{2r+2}$, and eigenvalue convergence of order $h^{r+2}$ under regularity, together with $E^0_h \le E$ and $O(h^2)$ convergence for the modified method.

What carries the argument

The load-bearing object is the pair (reconstruction, stabilisation): given cell unknowns in $P_{k+1}(T)$ and face unknowns in $P_k(F)$, the operator $R_h$ from (2.3) reconstructs a broken polynomial that matches the cell unknown's average and whose gradient against test polynomials equals the boundary and cell data, while $s_h$ penalises the mismatch between cell and face components with explicit weights. This defines the discrete energy $E_h$. For the lower bound, the modified energy $E^0_h$ changes only the quartic term to $\frac{\kappa}{4}(|\Pi^0_{T_h} v_{T_h}|^2 v_{T_h}, v_{T_h})$, and the proof chains a Poincaré inequality with constant $1/\pi^2$, a trace inequality with constant $C_{\mathrm{tr}}$, and Jensen's inequality to dominate each discrete term by the continuous energy. A moment-preserving smoother $J_h$ is used in the error analysis to convert nonconforming discrete errors into conforming test functions.

What would settle it

For a smooth potential, run the modified HHO method on refined meshes while keeping the piecewise-constant under-approximation of the potential on a fixed coarse grid, as in Section 5.1; if the gap $E-E^0_h$ does not shrink to zero as $h \to 0$ but levels off at a positive constant, then the claim of asymptotically exact lower bounds for the original smooth problem fails.

Watch

Extended reading notes

Core claim

The central discovery is that the HHO energy functional, minimised over the space $\hat{V}_h = P_{k+1}(T_h) \times P_k(F_h)$, reproduces for the nonlinear Gross–Pitaevskii problem the optimal convergence behaviour known for linear elliptic eigenvalue problems: the reconstructed ground state converges like $h^{r+1}$ in the H1-seminorm, the cell component like $h^{r+2}$ in L2, and the energy like $h^{2r+2}$; the eigenvalue error is $h^{r+2}$ under an additional regularity assumption. The lower-bound variant modifies only the nonlinear term, replacing $|v_{T_h}|^2 v_{T_h}$ by $|\Pi^0_{T_h} v_{T_h}|^2 v_{T_h}$, which makes Jensen's inequality applicable and yields $E^0_h \le E$ for piecewise-constant potentials under the explicit mesh condition (3.5). The same modified method is shown to converge at first order in H1 and second order in L2, energy, and eigenvalue, so the guaranteed lower bound is asymptotically exact without any post-processing step.

Load-bearing premise

The lower-bound theorem assumes the potential is constant on each mesh element and requires condition (3.5), which involves the unknown discrete energy, so in practice the guarantee depends on a separate upper-bound computation.

Editorial extensions

If this is right

  • A $k$-th order HHO discretisation computes a ground-state approximation with L2 error $O(h^{k+2})$ and H1 error $O(h^{k+1})$, and an energy error $O(h^{2k+2})$; under extra regularity the eigenvalue error is $O(h^{k+2})$.
  • The modified $k=0$ method produces an energy $E^0_h$ that is guaranteed to satisfy $E^0_h \le E$ and converges to $E$ at $O(h^2)$, so it can be combined with any upper-bound method to bracket the true ground-state energy.
  • The lower bound needs no post-processing, which numerical tests show makes it one to two orders of magnitude tighter than the post-processed lowest-order mixed method for smooth potentials.
  • For piecewise-constant potentials, such as the disorder-potential experiments, the guaranteed lower bound holds with explicit constants, and the convergence statements apply to the discrete solution produced by the method.
  • The analysis covers convex Lipschitz domains in two and three space dimensions with homogeneous Dirichlet conditions, the standard truncation model for confined condensates.

Reading between the lines

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

  • Inference: for smooth potentials the experiments under-approximate $V$ on a fixed coarse grid and never refine it, so the asymptotic exactness is proved for the modified potential; I would expect a fixed, non-vanishing gap $E-E^0_h$ as $h \to 0$ for the original smooth problem.
  • Inference: condition (3.5) contains the unknown discrete energy, so a self-contained method would need either an a-priori choice of $\sigma$ small enough or an upper bound computed by the same HHO scheme rather than an external conforming method.
  • Inference: because the lower bound is obtained before any post-processing, the same discrete solution could feed an adaptive loop that refines where the local gap between upper and lower energies is largest; the paper does not explore adaptivity.
  • Inference: the reliance on Jensen's inequality suggests that the proof technique itself limits guaranteed lower bounds to lowest order; a high-order guaranteed lower bound would need a different treatment of the nonlinearity.
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

3 major / 3 minor

Summary. The paper introduces a hybrid high-order (HHO) discretization for the Gross–Pitaevskii eigenvalue problem (GP-EVP). For the standard HHO method, it proves optimal-order a priori error estimates for the ground state, eigenvalue, and energy under regularity assumptions. It then introduces a lowest-order modified HHO method with a special quadrature for the nonlinearity and proves, under a piecewise-constant potential assumption and a condition involving the discrete energy, that the discrete energy is a guaranteed lower bound for the exact ground-state energy. The authors claim that, unlike previous mixed finite element approaches, these lower bounds are obtained directly without post-processing and are asymptotically exact. Numerical experiments for harmonic, lattice, and disorder potentials compare the modified HHO lower bounds with those of the post-processed Raviart–Thomas method from GHLP25, and convergence experiments illustrate the predicted rates.

Significance. If the main results hold, the paper makes a useful contribution: it extends the HHO methodology to a nonlinear eigenvalue problem, proves sharp-looking convergence rates for both the standard and modified methods, and provides a new avenue toward guaranteed lower energy bounds for the GP-EVP. The paper is careful in separating the convergence analysis for the unmodified HHO method from the lower-bound construction, and it supplies several explicit constants and auxiliary lemmas. The availability of reproducible code is a further strength. However, the headline claim of asymptotically exact lower energy bounds for smooth potentials is not supported by the theorems as stated: the numerical protocol freezes a coarse piecewise-constant under-approximation of the potential, so the computed lower bounds converge to the energy of a different potential. This is a scope issue with the main advertised novelty rather than an internal inconsistency in the convergence proofs.

major comments (3)
  1. [Abstract, Theorem 3.1, Remark 3.2, §5.1] The claim that the modified HHO method yields 'asymptotically exact lower energy bounds' is not established for the smooth potentials V1 and V2 used in the experiments. Theorem 3.1 requires V ∈ P0(Th). In §5.1 the authors satisfy this by replacing V1 and V2 with a piecewise-constant under-approximation on a fixed Cartesian grid of size h = 2^{-2} and prolonging it to all finer meshes. Since that potential is fixed, the modified method converges as h → 0 to E(V_coarse), the ground-state energy of the modified potential, not to E(V). The gap E(V) - E(V_coarse) is independent of h, so the plotted quantity E - E0_h cannot tend to zero. The lower-bound property is preserved because V_coarse ≤ V, but asymptotic exactness for the original problem is not. The abstract and conclusion should state this limitation explicitly, or the numerical protocol should be changed so that the potential under-approximation is refined with the mesh.
  2. [Theorem 3.1, Eq. (3.5), §5.1] Condition (3.5) is not a fully computable a priori certificate because it contains the unknown discrete energy E0_h. In §5.1 the authors make it hold by first computing an independent upper bound E with a conforming Q1 method and then choosing σ from (3.5) with E in place of E0_h. This is logically valid only after an external upper bound is available, and it makes the 'guaranteed' lower bound depend on a second computation. The paper should either state this dependence as an explicit hypothesis of Theorem 3.1, or provide a computable upper bound for E0_h that makes the guarantee self-contained. As written, the theorem guarantees E0_h ≤ E conditional on a condition that involves the very quantity to be bounded.
  3. [Theorem 4.8 and its proof] The convergence theorem for the modified HHO method, which underpins the second-order energy and eigenvalue claims, is not actually proved. The proof states that it is 'very similar' to the standard case and that the arguments of Lemmas 4.3-4.6 are repeated using Lemmas A.5-A.7, but the nontrivial adaptations—particularly the treatment of the modified nonlinear form (4.33) and the quadrature error estimates—are not carried out. Since the modified method is a central contribution, the proof should be written out in sufficient detail, or the theorem should be stated with a clear indication of which parts remain conjectural. This is especially important because the modified nonlinearity is non-polynomial in the discrete unknowns and the eigenvalue problem (4.33) requires separate handling.
minor comments (3)
  1. [Figure 5.3] The caption says 'polynomial degrees k = 1, 2, 3 (from left to right)', but the panels are labeled k = 0, 1, 2 and the text refers to k = 0, 1, 2; the caption should be corrected.
  2. [§5.1, disorder potential] The disorder potential V3 is described as piecewise constant on a Cartesian grid with mesh size h = 20; within Ω = (-8,8)^2 this value is ambiguous and is likely intended to be 2^0 or a similar small value. Please clarify.
  3. [Remark 3.2 and §5.1] Remark 3.2 suggests approximating V by its elementwise minimum, while §5.1 says the potentials are 'projected' onto piecewise constants. These are different procedures, and only an under-approximation preserves the lower-bound property. Please clarify which construction was actually used in the experiments.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular step found: the error analysis and lower-bound inequality are derived from standard HHO and GP results; only a non-load-bearing self-citation and a scope gap in the numerical lower-bound protocol prevent a pure 0.

full rationale

Walking the derivation chain, I find no step in which a prediction reduces to an input by construction. The convergence results (Theorems 4.2, 4.5, 4.6, 4.8) are built on classical HHO Poisson estimates [DPD20], the conforming GP analysis [CCM10], explicit Poincare/trace constants, and a two-term error splitting (interpolation error plus HHO approximation error); the target rates are not assumed. The lower-bound Theorem 3.1 is proved term-by-term via Pythagoras' identity, stability of L2-projections, and Jensen's inequality, yielding the explicit sufficient condition (3.5). Although (3.5) contains the unknown discrete energy E0_h, the paper uses an independent Q1 upper bound to choose sigma, so the inequality E0_h <= E is not fitted: it is a checkable a priori condition. The self-citations are real but non-load-bearing: [GHLP25] is used as a benchmark comparison, [HLP24] only in a remark on discrete uniqueness, and [LT25] merely provides a 'similar argument' for a smoother whose existence is cited to the independent paper [EZ20]. Two caveats are flagged for the reader but are not circularity. First, for smooth potentials V1/V2, Section 5.1 replaces the potential by a fixed coarse piecewise-constant under-approximation on a Cartesian grid with h=2^{-2} and prolongs it to all finer meshes; the computed lower bounds therefore converge to the energy of the modified potential, so asymptotic exactness for the original smooth problem is not established, and Theorem 3.1's assumption V in P0(Th) is not literally satisfied on the simplicial meshes. Second, Theorem 4.8's proof is mostly delegated ('the proof follows that of the standard HHO method... omitted'), an omission rather than a circular derivation. Score 2 reflects only the minor non-load-bearing self-citation; the central derivation is self-contained.

Assumptions & free parameters 2 free parameters · 7 assumptions · 0 invented entities

The analysis leans on standard HHO machinery, regularity assumptions, and a conditional inequality rather than on invented entities. The principal caveats are the coupling of the lower-bound guarantee to an external upper bound via (3.5), the piecewise-constant potential restriction in Theorem 3.1, and regularity assumptions on the exact ground state.

free parameters (2)
  • stabilization parameter sigma = order 1, chosen via (3.5) using a reference upper bound of E0_h
    User-chosen HHO penalty parameter; for guaranteed lower bounds it must satisfy (3.5), which in experiments is enforced by rearranging the inequality using an estimated upper bound of the exact energy.
  • piecewise-constant potential under-approximation grid size = h_pot = 2^{-2}
    For smooth potentials V1 and V2, the potential is replaced by its elementwise-minimum piecewise constant version on a fixed 2^{-2} grid; this hand-chosen modeling step creates a fixed bias in the guaranteed lower bound for the original problem.
assumptions (7)
  • domain assumption Existence, uniqueness (up to sign) of the GP ground state and simplicity of lambda1
    Relied on throughout (Section 1, cited to CCM10) to ensure the target is well-defined.
  • domain assumption Convexity of Omega and H2-regularity for auxiliary Poisson and dual problems
    Used in Theorems 4.2, 4.5, 4.6 and Lemma 4.4 to control the auxiliary problems and the smoothing operator.
  • domain assumption Regularity u in H^{r+2}(Th), and u^3 in H^m in Remark 4.7
    Assumed to state the convergence rates; not proved for the GP ground state in general.
  • domain assumption Condition (3.5): 1 - sigma(1/pi^2 + Ctr) - 4 h^2 E0_h / pi^2 >= 0
    The guaranteed lower-bound theorem is conditional on this inequality involving sigma, h, and the unknown discrete energy E0_h; it is verified a posteriori in Section 5.1 using an external upper bound.
  • domain assumption The discrete minimizer exists and is actually found by the solver
    Section 2 states 'we proceed under the assumption that a discrete minimiser has been found'; existence holds by compactness, but global convergence of the chosen gradient flow is not proved.
  • standard math Known HHO approximation properties, moment-preserving smoother, and discrete Sobolev embeddings
    Cited from DPD20, EZ20, and DPD17 and used as black boxes in Lemmas 4.3, 4.4 and A.3-A.4.
  • standard math Poincare and trace inequalities with explicit constants
    Lemmas A.1-A.2, cited to Beb03 and Tra24/Gal23, are used in the lower-bound proof.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Hybrid High-Order Method for the Gross--Pitaevskii Eigenvalue Problem." pith.science (2026). https://pith.science/paper/YLDKK6QZ

@misc{pith2026250619944,
  author       = {Pith},
  title        = {Pith review of: A Hybrid High-Order Method for the Gross--Pitaevskii Eigenvalue Problem},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YLDKK6QZ}},
  note         = {Machine review of arXiv:2506.19944}
}
read the original abstract

We introduce a hybrid high-order method for approximating the ground state of the nonlinear Gross--Pitaevskii eigenvalue problem. Optimal convergence rates are proved for the ground state approximation, as well as for the associated eigenvalue and energy approximations. Unlike classical conforming methods, which inherently provide upper bounds on the ground state energy, the proposed approach gives rise to guaranteed and asymptotically exact lower energy bounds. Importantly, and in contrast to previous works, they are obtained directly without the need of post-processing, leading to more accurate guaranteed lower energy bounds in practice.

Figures

Figures reproduced from arXiv: 2506.19944 by the authors.

Figure 5.1
Figure 5.1. First row: Potentials V1, V2, and V3 (from left to right). Second row: corresponding ground state approximations. values on each grid element are assigned randomly as independent realisations of coin-toss variables taking values in {10, 50}. The particle interaction parameter is set to κ = 1. Illustrations of the potentials and corresponding ground state approximations are provided in [PITH_FULL_IMAGE:figures/full_… view at source ↗
Figure 5.2
Figure 5.2. First row: energy approximations E0 h of the modified HHO method and E RT0 h of the post-processed Raviart–Thomas method (see [GHLP25]) for the potentials V1, V2, and V3 (from left to right). Second row: corresponding approximation errors. 5.2. Optimal order convergence. In the following, we investigate the conver￾gence properties of the HHO method introduced in (2.4) and its modified version from (3.2). For the cor… view at source ↗
Figure 5.3
Figure 5.3. First row: convergence plots of the (reconstructed) ground state approximations of the HHO method for polynomial degrees k = 1, 2, 3 (from left to right). Second row: convergence plots of the corresponding eigenvalue and energy approximations. 2−6 2−5 2−4 2−3 2−2 2−1 2 0 10−7 10−6 10−5 10−4 10−3 10−2 10−1 100 101 h 7→ ∥u − uh∥L2(Ω) h 7→ ∥u − Ruh∥L2(Ω) h 7→ ∥∇(u − Ruh)∥L2(Ω) h 7→ 0.2h 1 h 7→ 0.01h 2 2−6 2−5 2−4 2−3 2… view at source ↗
Figures from the paper (1 more)
Figure 5.4
Figure 5.4. Figure 5.4: Left: convergence plot of the ground state approxi￾mation of the modified HHO method. Right: convergence plot of the corresponding eigenvalue and energy approximations (right). any post-processing. Numerical experiments confirm that, particularly for smooth problems,…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 31 canonical work pages

  1. [1]

    Alberty, C

    J. Alberty, C. Carstensen, and S. Funken, Remarks around 50 lines of M atlab: short finite element implementation , Numer. Algorithms 20 (1999), no. 2--3, 117--137. 1709562

    work page 1999

  2. [2]

    Altmann, P

    R. Altmann, P. Henning, and D. Peterseim, Quantitative A nderson localization of S chr\"odinger eigenstates under disorder potentials , Math. Models Methods Appl. Sci. 30 (2020), no. 5, 917--955

    work page 2020

  3. [3]

    , Localization and delocalization of ground states of B ose-- E instein condensates under disorder , SIAM J. Appl. Math. 82 (2022), no. 1, 330--358

    work page 2022

  4. [4]

    J. R. Abo-Shaeer, C. Raman, J. M. Vogels, and W. Ketterle, Observation of vortex lattices in B ose-- E instein condensates , Science 292 (2001), no. 5516, 476–479

    work page 2001

  5. [5]

    Bao and Y

    W. Bao and Y. Cai, Mathematical theory and numerical methods for B ose-- E instein condensation , Kinet. Relat. Models 6 (2013), no. 1, 1--135

    work page 2013

  6. [6]

    Bebendorf, A note on the poincaré inequality for convex domains, Z

    M. Bebendorf, A note on the poincaré inequality for convex domains, Z. Anal. Anwend. 22 (2003), no. 4, 751–756

    work page 2003

  7. [7]

    Canc \`e s, R

    E. Canc \`e s, R. Chakir, and Y. Maday, Numerical analysis of nonlinear eigenvalue problems, J. Sci. Comput. 45 (2010), no. 1-3, 90--117

    work page 2010

  8. [8]

    Carstensen, A

    C. Carstensen, A. Ern, and S. Puttkammer, Guaranteed lower bounds on eigenvalues of elliptic operators with a hybrid high-order method, Numer. Math. 149 (2021), no. 2, 273–304

    work page 2021

Show all 33 references
  1. [9]

    Carstensen, B

    C. Carstensen, B. Gr\" a ßle, and N. T. Tran, Adaptive hybrid high-order method for guaranteed lower eigenvalue bounds, Numer. Math. 156 (2024), no. 3, 813–851

    2024

  2. [10]

    H. Chen, L. He, and A. Zhou, Finite element approximations of nonlinear eigenvalue problems in quantum physics, Comp. Meth. Appl. Mech. Eng. 200 (2011), no. 21-22, 1846--1865

    2011

  3. [11]

    Z. Chen, J. Lu, Y. Lu, and X. Zhang, Fully discretized S obolev gradient flow for the G ross-- P itaevskii eigenvalue problem , ArXiv e-print 2403.06028 (2024)

    2024 arXiv

  4. [12]

    Carstensen, Q

    C. Carstensen, Q. Zhai, and R. Zhang, A skeletal finite element method can compute lower eigenvalue bounds, SIAM J. Numer. Anal. 58 (2020), no. 1, 109–124

    2020

  5. [13]

    Dalfovo, S

    F. Dalfovo, S. Giorgini, L. P. Pitaevskii, and S. Stringari, Theory of B ose-- E instein condensation in trapped gases , Rev. Mod. Phys. 71 (1999), no. 3, 463–512

    1999

  6. [14]

    Danaila and F

    I. Danaila and F. Hecht, A finite element method with mesh adaptivity for computing vortex states in fast-rotating B ose-- E instein condensates , J. Comput. Phys. 229 (2010), no. 19, 6946--6960

    2010

  7. [15]

    Di Pietro and J

    D. Di Pietro and J. Droniou, A hybrid high-order method for L eray- L ions elliptic equations on general meshes , Math. Comp. 86 (2017), no. 307, 2159--2191. 3647954

    2017

  8. [16]

    Modeling, Simulation and Applications, vol

    , The hybrid high-order method for polytopal meshes, MS&A. Modeling, Simulation and Applications, vol. 19, Springer, Cham, 2020. 4230986

    2020

  9. [17]

    Di Pietro and R

    D. Di Pietro and R. Tittarelli, An introduction to hybrid high-order methods, p. 75–128, Springer International Publishing, 2018

    2018

  10. [18]

    Ern and P

    A. Ern and P. Zanotti, A quasi-optimal variant of the hybrid high-order method for elliptic partial differential equations with H^ -1 loads , IMA J. Numer. Anal. 40 (2020), no. 4, 2163--2188. 4167044

    2020

  11. [19]

    Fetter, Rotating trapped B ose-- E instein condensates , Rev

    A. Fetter, Rotating trapped B ose-- E instein condensates , Rev. Mod. Phys. 81 (2009), no. 2, 647–691

    2009

  12. [20]

    Gallistl, Mixed methods and lower eigenvalue bounds, Math

    D. Gallistl, Mixed methods and lower eigenvalue bounds, Math. Comp. 92 (2023), no. 342, 1491--1509

    2023

  13. [21]

    Gallistl, M

    D. Gallistl, M. Hauck, Y. Liang, and D. Peterseim, Mixed finite elements for the G ross-- P itaevskii eigenvalue problem: a priori error analysis and guaranteed lower energy bound , IMA J. Numer. Anal. 45 (2025), no. 3, 1320--1346. 4917676

    2025

  14. [22]

    Henning and E

    P. Henning and E. Jarlebring, The G ross-- P itaevskii E quation and E igenvector N onlinearities: N umerical M ethods and A lgorithms , SIAM Rev. 67 (2025), no. 2, 256--317

    2025

  15. [23]

    Hauck, Y

    M. Hauck, Y. Liang, and D. Peterseim, Positivity preserving finite element method for the G ross-- P itaevskii ground state: discrete uniqueness and global convergence , ArXiv e-print 2405.17090 (2024)

    2024 arXiv

  16. [24]

    Henning, A

    P. Henning, A. M lqvist, and D. Peterseim, Two-level discretization techniques for ground state computations of B ose-- E instein condensates , SIAM J. Numer. Anal. 52 (2014), no. 4, 1525--1550

    2014

  17. [25]

    Henning and D

    P. Henning and D. Peterseim, Sobolev gradient flow for the G ross-- P itaevskii eigenvalue problem: global convergence and computational efficiency , SIAM J. Numer. Anal. 58 (2020), no. 3, 1744--1772

    2020

  18. [26]

    Henning and A

    P. Henning and A. Persson, On optimal convergence rates for discrete minimizers of the G ross-- P itaevskii energy in localized orthogonal decomposition spaces , Multiscale Model. Simul. 21 (2023), no. 3, 993–1011

    2023

  19. [27]

    P. Heid, B. Stamm, and T. P. Wihler, Gradient flow finite element discretizations with energy-based adaptivity for the G ross-- P itaevskii equation , J. Comput. Phys. 436 (2021), Paper No. 110165, 15

    2021

  20. [28]

    Henning and J

    P. Henning and J. W\" a rneg rd, Superconvergence of time invariants for the G ross-- P itaevskii equation , Math. Comp. 91 (2022), no. 334, 509--555

    2022

  21. [29]

    Henning and M

    P. Henning and M. Yadav, On discrete ground states of rotating B ose-- E instein condensates , 2024

    2024

  22. [30]

    Liang and N

    Y. Liang and N. T. Tran, A hybrid high-order method for the biharmonic problem , arXiv e-prints (2025), arXiv:2504.16608

    2025 arXiv

  23. [31]

    Peterseim, J

    D. Peterseim, J. Wärnegård, and C. Zimmer, Super-localised wave function approximation of B ose-- E instein condensates , J. Comput. Phys. 510 (2024), 113097

    2024

  24. [32]

    N. T. Tran, Lower eigenvalue bounds with hybrid high-order methods , 2024, p. arXiv:2406.06244

    2024 arXiv

  25. [33]

    Zhou, An analysis of finite-dimensional approximations for the ground state solution of B ose-- E instein condensates , Nonlinearity 17 (2004), no

    A. Zhou, An analysis of finite-dimensional approximations for the ground state solution of B ose-- E instein condensates , Nonlinearity 17 (2004), no. 2, 541--550

    2004

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.