Pith. sign in

REVIEW 3 major objections 5 minor 73 references

Chaoticus: a parallel approach to the computation of chaos indicators

T0 review · 3 major / 5 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 Chaoticus is a GPU-accelerated Python package for computing SALI, GALI, Lagrangian-descriptor, and Lyapunov-spectrum chaos indicators, claiming speedups of several orders of magnitude over CPU solvers.

desk verdict A plausible GPU chaos-indicator package, but the speedup claim and even the code itself are not verifiable from this preprint. read the letter →

arxiv 2507.00622 v1 pith:XL2BKXGX submitted 2025-07-01 nlin.CD math.DSphysics.comp-ph

Javier Jiménez-López , José Sáez-Landete , Victor J. García-Garrido This is my paper · ORCID
classification nlin.CDmath.DSphysics.comp-ph
keywords chaosindicatorsHamiltoniandynamicsLagrangiandescriptorsGPUaccelerationSALIGALILyapunovexponentsparallelcomputing
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 presents Chaoticus, a Python package for GPU-accelerated numerical integration of Hamiltonian ODE systems and the computation of chaos indicators: SALI, GALI, Lagrangian-descriptor-based indicators, and the Lyapunov exponent spectrum. Its central claim is that GPU parallelization reduces computation times by several orders of magnitude compared to CPU-based solvers, while preserving energy conservation at the $10^{-8}$--$10^{-9}$ level. Such speedups would make large-scale scans of phase space---classifying millions of trajectories as regular or chaotic---practicable for systems like the double pendulum, H\'enon--Heiles, and Fermi--Pasta--Ulam. The paper is a software contribution: it benchmarks the package on three Hamiltonian systems to show that integration time per initial condition drops as more trajectories are packed onto the GPU until its memory saturates.

What carries the argument

The load-bearing mechanism is a set of GPU kernels implementing the DOP8 embedded Runge--Kutta method (a Dormand--Prince 8(5) scheme) in fixed-step and adaptive-step variants, which integrate many trajectories in parallel. Around these sit kernels that normalize deviation vectors for SALI/GALI, perform QR factorization for the Lyapunov spectrum, and compute SVD for the GALI indicator; Lagrangian-descriptor indicators are accumulated from the trajectory itself without needing variational equations. The package also tracks per-step error for the adaptive solvers and reports energy drift. The effects of the machinery are shown in figures where the integration time per initial condition decreases as the number of simultaneously integrated conditions increases up to GPU saturation, and where the GPU time grows roughly linearly with the number of degrees of freedom for the FPU system, unlike the CPU baseline.

What would settle it

Run the FPU/GALI4 benchmark with $10^3$ fixed random initial conditions, integration time $\tau = 10^3$, and an absolute tolerance of $10^{-8}$ for both the CPU DOP853 reference solver and the GPU kernel on the same hardware, then compare wall-clock times at matched energy drift. If the GPU version is not several orders of magnitude faster at equal $\Delta H \sim 10^{-8}$, the central speedup claim fails.

Watch

Extended reading notes

Core claim

The central discovery is that a well-parallelized GPU implementation of an 8th-order Dormand--Prince integrator, together with companion kernels that evolve deviation vectors and perform SVD or QR decompositions, can compute standard chaos indicators for large ensembles of Hamiltonian trajectories at a fraction of the CPU time. Concretely, Chaoticus integrates thousands of initial conditions simultaneously on the GPU, with fixed- or adaptive-step solvers, and the wall-clock time per trajectory falls as the GPU's local memory is filled. The package is validated on the dimensionless double pendulum, the H\'enon--Heiles system, and the Fermi--Pasta--Ulam chain, with energy conservation of order $10^{-9}$ for the double pendulum, $10^{-8}$ for H\'enon--Heiles, and at least $10^{-8}$ for the FPU/GALI4 comparison against the CPU DOP853 baseline. The authors conclude that this reduction in computing time enables the generation of extensive datasets for in-depth analysis of complex dynamics in Hamiltonian systems.

Load-bearing premise

The claim of orders-of-magnitude speedup presumes that the GPU and CPU solvers are compared at the same numerical accuracy and with equivalent step-size control, so that the observed time reduction is due to parallelism rather than looser tolerances or different problem setups.

Editorial extensions

If this is right

  • Phase-space cartography of Hamiltonian systems at high resolution becomes affordable: scans with millions of initial conditions for SALI or GALI can be produced in the time previously needed for thousands.
  • Lagrangian-descriptor-based chaos indicators, which need no variational equations, can be computed for large ensembles on GPUs, simplifying high-throughput chaos detection.
  • The fixed-step and adaptive-step solvers let users trade raw speed against robustness for highly nonlinear systems, both conserving energy to $10^{-8}$ or better in the test cases.
  • The GPU approach scales near-linearly with the number of degrees of freedom in the FPU test, so high-dimensional Hamiltonian lattices can be probed at scales that are impractical on CPU.
  • Because the package implements the full pipeline in Python, researchers can generate extensive labeled datasets of regular versus chaotic trajectories, which could serve as training data for machine-learning models of dynamical behavior.

Reading between the lines

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

  • The 'orders of magnitude' speedup is demonstrated for batch ensembles that saturate the GPU; for a single trajectory or a small ensemble, the CPU could well be faster or comparable, so the benefit is inherently throughput-oriented.
  • The same GPU-parallel pattern could be applied to indicators not included here, such as MEGNO or fast Lyapunov indicators, and to symplectic integrators, which might preserve energy even more faithfully over long integration times.
  • A rigorous benchmark would report time-to-solution at matched error (e.g., equal energy drift or equal local truncation error) and identify the GPU/CPU hardware, which would make the speedup quantitative rather than indicative.
  • The package's approach could be extended to non-Hamiltonian ODE systems, since the integrator and indicator kernels are not inherently restricted to conservative dynamics.
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

3 major / 5 minor

Summary. The manuscript introduces Chaoticus, a Python package for GPU-accelerated integration of Hamiltonian ODE systems and computation of chaos indicators including SALI, GALI, Lagrangian-descriptor-based indicators, and Lyapunov spectra. It reports fixed-step and adaptive DOP8 solvers with variational-equation support, and presents timing experiments on the double pendulum, Hénon-Heiles, and Fermi-Pasta-Ulam systems. The central claim is that GPU parallelization reduces computation times by several orders of magnitude relative to CPU implementations.

Significance. If the speedup and accuracy claims hold, Chaoticus would be a genuinely useful contribution: SALI/GALI and Lagrangian descriptors are standard but computationally expensive, and GPU batching across initial conditions is a natural and potentially high-impact optimization. The paper relies on established indicators and solvers, with no circularity in the definitions of the computed quantities. However, the evidence actually presented is currently insufficient: the benchmark setup is underspecified, no code or hardware details are given, and the accuracy validation does not exercise the quantities being computed. The significance is therefore prospective rather than demonstrated.

major comments (3)
  1. [Section 4, Fig. 3; Code availability] The abstract's claim of 'several orders of magnitude' speedup is not supported by the information given. Fig. 3 reports a comparison between SciPy's DOP853 (with absolute tolerance 1e-8) and 'our implementation in GPU', but the manuscript does not specify the GPU hardware, the CPU model, whether the GPU used the fixed-step DOP8 or adaptive DOP8(5) kernel, the number of repeated timings, the run-to-run variance, or how the 10^3 CPU trajectories were scheduled (e.g., loop versus vectorized). Without these details, and without the code (the Code availability section gives no URL, version, or commit), the timing comparison cannot be reproduced or independently evaluated. This is load-bearing because the speedup is the paper's central contribution.
  2. [Section 4, Figs. 1-3] Energy conservation alone does not validate the computed chaos indicators. The reported ΔH ~ 1e-8 to 1e-9 checks the Hamiltonian flow, but SALI/GALI depend on the evolution of deviation vectors through the variational equations; a fixed-step and an adaptive solver can both conserve energy to similar precision while producing different phase-space trajectories and different alignment indices, especially for chaotic orbits over τ ~ 10^3 to 10^4. No comparison against known SALI/GALI values, published reference results, or an independent CPU implementation of the same indicators is provided. The manuscript should show, for example, that GALI4 for representative FPU initial conditions converges to the known values (near zero for chaotic orbits and nonzero for regular ones) and agrees with a reference integrator.
  3. [Section 3 and Section 4] The paper advertises many functionalities (QR-based Lyapunov spectra, per-step error computation, neighboring-trajectory generation) but provides no demonstration or tests for most of them; only timing curves for a subset appear in Section 4. For a software paper, a minimal artifact such as a repository with test cases and a benchmark script is needed to substantiate that these functions work. The phrase 'can be found in Chaoticus' in the Code availability section is not a locator, and the absence of a URL, version, or DOI makes the package effectively inaccessible to the reader.
minor comments (5)
  1. [Section 2] The name 'Bikhoff averages' should be 'Birkhoff averages'.
  2. [Section 4] The text 'setted' should be 'set', and the Fig. 3 caption text contains 'liner trend' which should be 'linear trend'.
  3. [Figure captions] The figure captions state energy conservation at ΔH ~ 1e-8 or 1e-9 but do not define how ΔH is computed or whether it is a maximum, mean, or final error; this should be specified.
  4. [Equations] Several equations have formatting artifacts (for example, the matrix expressions in Eq. (3) and the inline fraction in Eq. (6)); the final version should be carefully typeset.
  5. [Code availability] The repository link is missing; the authors should provide a URL, version, and license, and should cite their own software in the references.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper reports a GPU-accelerated software package and empirical speedups; its chaos indicators and integrators are standard published methods, and no fit or self-citation defines the claimed result.

full rationale

The manuscript is a software-presentation paper rather than a derivation-based study. The chaos indicators (SALI, GALI, Lagrangian-descriptor indicators, Lyapunov spectra) are implemented according to established published definitions, cited to the original literature (e.g., [54, 56, 57] for SALI, [58] for GALI, [41, 38] for Lagrangian descriptors), and the ODE integrators are based on the published Dormand-Prince DOP853 scheme [51]. No quantity is defined in terms of another quantity that it is then said to predict; there is no fitted parameter that is later renamed as a prediction. The central claim of the paper is the empirical GPU speedup, which is presented as a measured comparison in Figs. 1-3 rather than as a derived result. The self-citations ([29], [30], [31]) point to prior work on Lagrangian descriptors and double-pendulum dynamics, but none of these is used as the justification for the package's correctness or for the speedup claim; they are related-work references and do not carry the argument. The main weakness of the paper is that the benchmark details (GPU/CPU hardware, solver variants, timing methodology, code repository URL) are not reported, which affects verifiability and reproducibility, but this is a completeness and rigor issue, not circularity. The 'Code availability' section merely states that the library 'can be found in Chaoticus' without a URL or version, which is also a reproducibility gap rather than a circular reasoning pattern. Applying the specified criteria, there is no step in which a prediction reduces by construction to an input, no self-citation chain that forces the outcome, and no ansatz smuggled in via citation. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper is an engineering contribution and introduces no new mathematical postulates, fitted parameters, or new physical entities. The only assumptions are domain-standard modeling choices and trust in the numerical methods, which the paper does not independently validate.

assumptions (3)
  • domain assumption The published Hamiltonians for the double pendulum, Hénon-Heiles, and Fermi-Pasta-Ulam systems correctly encode the dynamics of those systems.
    Section 4 uses these models as test cases, and the paper does not discuss any modification or alternative formulation.
  • domain assumption The DOP853 integrator with the stated tolerances produces sufficiently accurate trajectories and chaos indicator values.
    The only accuracy check reported is energy conservation at a level of 1e-8 to 1e-9; no validation of the actual SALI, GALI, or Lyapunov values against known analytic or reference results is provided.
  • domain assumption GPU kernels return numerically equivalent results to the CPU integrator for the same inputs and step sizes.
    The paper assumes the GPU implementation is a faithful port of DOP853 and the indicator algorithms, but provides no comparison of indicator outputs between CPU and GPU runs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Chaoticus: a parallel approach to the computation of chaos indicators." pith.science (2026). https://pith.science/paper/XL2BKXGX

@misc{pith2026250700622,
  author       = {Pith},
  title        = {Pith review of: Chaoticus: a parallel approach to the computation of chaos indicators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XL2BKXGX}},
  note         = {Machine review of arXiv:2507.00622}
}
read the original abstract

In this paper we present Chaoticus, a Python-based package for the GPU-accelerated integration of ODE systems and the computation of chaos indicators, including SALI, GALI, Lagrangian Descriptors based indicators and the Lyapunov exponent spectrum. By leveraging GPU parallelization, our package significantly reduces the computation times by several orders of magnitude compared to CPU-based approaches. This significant reduction in computing time facilitates the generation of extensive datasets, crucial for the in-depth analysis of complex dynamics in Hamiltonian systems.

Figures

Figures reproduced from arXiv: 2507.00622 by the authors.

Figure 1
Figure 1. Integration time per initial condition as a function of the number of simultaneously [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Integration time per initial condition as a function of the number of simultaneously [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Integration time as a function of the amount of DoF considered for the FPU system [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

73 extracted references · 58 canonical work pages

  1. [1]

    Springer Science & Busi- ness Media, 2001

    Fundamentals of astrodynamics and applications , volume 12. Springer Science & Busi- ness Media, 2001

    work page 2001

  2. [2]

    Aguilar-Sanjuan, V

    B. Aguilar-Sanjuan, V. J. García-Garrido, V. Krajňák, S. Naik, and S. Wiggins. Ldds: Python package for computing and visualizing lagrangian descriptors for dynamical systems. Journal of Open Source Software , 6(65):3482, 2021. doi:10.21105/joss. 03482

    doi:10.21105/joss 2021

  3. [3]

    A. A. Amarot Yagbem, M. V. Tchakui, H. Simo, and P. Woafo. Maximal Lyapunov Exponent and Smaller Alignment Index Computation Characterization of an Electro- dynamic Electromechanical System Powered by Sine, Square and Triangle Waves Elec- trical Signals. International Journal of Bifurcation and Chaos , 35(04):2550042, 2025. doi:10.1142/S0218127425500427. 6

    work page doi:10.1142/s0218127425500427 2025

  4. [4]

    A first taste of nonlinear beam dynamics

    H.Bartosik, Y.Papaphilippou, andA.Wolski. Afirsttasteofnonlinearbeamdynamics. In CAS - CERN Accelerator School 2021: Introduction to Accelerator Physics , 1 2022. arXiv:2201.01532

    work page Pith review arXiv 2021

  5. [5]

    R. R. Bate, D. D. Mueller, J. E. White, and W. W. Saylor.Fundamentals of astrody- namics. Courier Dover Publications, 2020

    work page 2020

  6. [6]

    Z. Chen, J. Zhang, M. Arjovsky, and L. Bottou. Symplectic recurrent neural networks,

    work page

  7. [7]

    P. M. Cincotta and C. Simó. Simple tools to study global dynamics in non-axisymmetric galactic potentials – i. Astron. Astrophys. Suppl. Ser. , 147(2):205–228, 2000. doi: 10.1051/aas:2000108

    doi:10.1051/aas:2000108

  8. [8]

    Colagrossi, S

    A. Colagrossi, S. Lizy-Destrez, N. Baresi, J. Masdemont, and L. Bucci. Astrodynamics, guidance, navigation and control in chaotic multi-body environments, 2022

    work page 2022

Show all 73 references
  1. [9]

    G. T. Craven, A. Junginger, and R. Hernandez. Lagrangian descriptors of driven chem- ical reaction manifolds.Phys. Rev. E, 96:022222, Aug 2017. URL:https://link.aps. org/doi/10.1103/PhysRevE.96.022222, doi:10.1103/PhysRevE.96.022222

    2017 doi

  2. [10]

    Căliman, J

    A. Căliman, J. Daquin, and A.-S. Libert. Improved detection of chaos with la- grangian descriptors using differential algebra. Physica D: Nonlinear Phenomena , 472:134506, 2025. URL: https://www.sciencedirect.com/science/article/pii/ S0167278924004561, doi:10.1016/j.physd.2024.134506

    2025

  3. [11]

    Daquin, R

    J. Daquin, R. Pédenon-Orlanducci, M. Agaoglou, G. García-Sánchez, and A. M. Man- cho. Global dynamics visualisation from lagrangian descriptors. applications to dis- crete and continuous systems. Physica D: Nonlinear Phenomena , 442:133520, 2022. doi:10.1016/j.physd.2022.133520

    2022

  4. [12]

    Measuring quasiperiodicity

    S.Das, C.Dock, Y.Saiki, M.Salgado-Flores, E.Sander, J.Wu, andJ.Yorke. Measuring quasiperiodicity. Europhysics Letters, 114(4):40005, 2016. doi:10.1209/0295-5075/ 114/40005

    2016 doi

  5. [13]

    David and F

    M. David and F. Méhats. Symplectic learning for Hamiltonian neural networks.Journal of Computational Physics , 494:112495, 2023. doi:10.1016/j.jcp.2023.112495

    2023

  6. [14]

    Dietz, A

    B. Dietz, A. Heusler, K. H. Maier, A. Richter, and B. A. Brown. Chaos and Regularity in the Doubly Magic Nucleus208Pb. Phys. Rev. Lett., 118:012501, 2017.doi:10.1103/ PhysRevLett.118.012501

    2017

  7. [15]

    Dvorak, F

    R. Dvorak, F. Freistetter, and J. Kurths.Chaos and Stability in Planetary Systems , volume 683. 2005. doi:10.1007/b94975

    2005 doi

  8. [16]

    Eleuch and A

    H. Eleuch and A. Prasad. Chaos and regularity in semiconductor microcavities.Physics Letters A, 376(26):1970–1977, 2012. doi:10.1016/j.physleta.2012.04.050

    1970 doi

  9. [17]

    Fermi, P

    E. Fermi, P. Pasta, S. Ulam, and M. Tsingou. Studies of the nonlinear problems. Techni- cal report, Los Alamos National Laboratory (LANL), Los Alamos, NM (United States), 05 1955. URL: https://www.osti.gov/biblio/4376203, doi:10.2172/4376203

    1955

  10. [18]

    Froeschlé, R

    C. Froeschlé, R. Gonczi, and E. Lega. The fast Lyapunov indicator: a simple tool to detect weak chaos. Application to the structure of the main asteroidal belt.Planetary and Space Science, 45(7):881–886, 1997. doi:10.1016/S0032-0633(97)00058-5

    1997 doi

  11. [19]

    Froeschlé, E

    C. Froeschlé, E. Lega, and R. Gonczi. Fast Lyapunov indicators. Application to as- teroidal motion. Celestial Mechanics and Dynamical Astronomy , 67(1):41–62, 1997. doi:10.1023/A:1008276418601

    1997 doi

  12. [20]

    P. Gaspard. Microscopic chaos and chemical reactions.Physica A: Statistical Mechanics and its Applications , 263(1):315–328, 1999. Proceedings of the 20th IUPAP Interna- tional Conference on Statistical Physics. URL: https://www.sciencedirect.com/ science/article/pii/S03784371...

    1999 doi

  13. [21]

    Gkolias, J

    I. Gkolias, J. Daquin, D. K. Skoulidou, K. Tsiganis, and C. Efthymiopoulos. Chaotic transport of navigation satellites. Chaos: An Interdisciplinary Journal of Nonlinear Science, 29(10):101106, 2019. doi:10.1063/1.5124682

    2019 doi

  14. [22]

    Greydanus, M

    S. Greydanus, M. Dzamba, and J. Yosinski. Hamiltonian Neural Networks. In H. Wal- lach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. Fox, and R. Garnett, editors, Advances in Neural Information Processing Systems , volume 32. Curran Associates, Inc., 2019

    2019

  15. [23]

    C.-D. Han, B. Glaz, M. Haile, and Y.-C. Lai. Adaptable hamiltonian neural networks. Phys. Rev. Res. , 3:023156, May 2021. URL: https://link.aps.org/doi/10.1103/ PhysRevResearch.3.023156, doi:10.1103/PhysRevResearch.3.023156

    2021 doi

  16. [24]

    Hénon and C

    M. Hénon and C. Heiles. The Applicability of the Third Integral of Motion: Some Numerical Experiments. Astronomical Journal, Vol. 69, p. 73 (1964) , 69:73, 1964. doi:10.1086/109234

    1964 doi

  17. [25]

    Hillebrand, G

    M. Hillebrand, G. Kalosakas, A. Schwellnus, and C. Skokos. Heterogeneity and chaos in the Peyrard-Bishop-Dauxois DNA model. Phys. Rev. E , 99:022213, 2019. doi: 10.1103/PhysRevE.99.022213

    2019 doi

  18. [26]

    Hillebrand, S

    M. Hillebrand, S. Zimper, A. Ngapasare, M. Katsanikas, S. Wiggins, and C. Skokos. Quantifying chaos using lagrangian descriptors. Chaos: An Interdisciplinary Journal of Nonlinear Science, 32(12):123122, 2022. doi:10.1063/5.0120889

    2022 doi

  19. [27]

    Hwang, C

    K. Hwang, C. Mitchell, and R. Ryne. Chaos Indicators for Studying Dynamic Aperture in the IOTA Ring with Protons. In10th International Particle Accelerator Conference, page WEPTS078, 2019. doi:10.18429/JACoW-IPAC2019-WEPTS078

    2019 doi

  20. [28]

    FromordertochaosinEarth satellite orbits

    F.G.IoannisGkolias, JérômeDaquinandA.J.Rosengren. FromordertochaosinEarth satellite orbits. The Astronomical Journal , 152(5), 2016. doi:10.3847/0004-6256/ 152/5/119

    2016 doi

  21. [29]

    Jiménez-López and V

    J. Jiménez-López and V. J. García-Garrido. Chaos and regularity in the double pen- dulum with lagrangian descriptors. International Journal of Bifurcation and Chaos , 34(16):2450201, 2024. arXiv:https://doi.org/10.1142/S0218127424502018, doi: 10.1142/S0218127424502018

    2024 doi

  22. [30]

    Jiménez-López and V

    J. Jiménez-López and V. García-Garrido. Learning the chaotic and regular nature of trajectoriesinhamiltoniansystemswithlagrangiandescriptors. Chaos, Solitons & Frac- tals, 191:115876, 2025. URL: https://www.sciencedirect.com/science/article/ pii/S0960077924014280, doi:10.1016/...

    2025

  23. [31]

    Jiménez-López and V

    J. Jiménez-López and V. J. García-Garrido. The simplest chaos indicator derived from lagrangian descriptors, 2025. URL: https://arxiv.org/abs/2506.16660, arXiv: 2506.16660

    2025

  24. [32]

    Junginger and R

    A. Junginger and R. Hernandez. Uncovering the geometry of barrierless reactions using lagrangian descriptors. The Journal of Physical Chemistry B , 120(8):1720–1725, Mar

  25. [33]

    Katsanikas, V

    M. Katsanikas, V. J. García-Garrido, and S. Wiggins. Detection of dynamical matching in a caldera hamiltonian system using lagrangian descriptors. International Journal of Bifurcation and Chaos , 30(09):2030026, 2020. arXiv:https://doi.org/10.1142/ S0218127420300268, doi:10.11...

    2020 doi

  26. [34]

    J. Laskar. Frequency analysis for multi-dimensional systems. global dynamics and diffusion. Physica D: Nonlinear Phenomena , 67(1):257–281, 1993. doi:10.1016/ 0167-2789(93)90210-R

    1993

  27. [35]

    Levnajić and I

    Z. Levnajić and I. Mezić. Ergodic theory and visualization. I. Mesochronic plots for vi- sualization of ergodic partition and invariant sets.Chaos: An Interdisciplinary Journal of Nonlinear Science, 20(3):033114, 2010. doi:10.1063/1.3458896. 8

    2010 doi

  28. [36]

    Levnajić and I

    Z. Levnajić and I. Mezić. Ergodic theory and visualization. II. Fourier mesochronic plots visualize (quasi)periodic sets. Chaos: An Interdisciplinary Journal of Nonlinear Science, 25(5):053105, 2015. doi:10.1063/1.4919767

    2015 doi

  29. [37]

    Y. Li, J. Wan, A. Liu, Y. Jiao, and R. Rainer. Data-driven chaos indicator for nonlinear dynamics and applications on storage ring lattice design. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 10...

    2022

  30. [38]

    Lopesino, F

    C. Lopesino, F. Balibrea-Iniesta, V. J. García-Garrido, S. Wiggins, and A. M. Mancho. A theoretical framework for lagrangian descriptors.International Journal of Bifurcation and Chaos, 27:1730001, 2017. doi:10.1142/S0218127417300014

    2017 doi

  31. [39]

    E. E. Macau and C. Grebogi. Control of chaos and its relevancy to spacecraft steer- ing. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 364(1846):2463–2481, 2006

    2006

  32. [40]

    Macek and A

    M. Macek and A. Leviatan. Regularity and chaos at critical points of first-order quan- tum phase transitions. Phys. Rev. C , 84:041302, 2011. doi:10.1103/PhysRevC.84. 041302

    2011 doi

  33. [41]

    A. M. Mancho, S. Wiggins, J. Curbelo, and C. Mendoza. Lagrangian descriptors: A method for revealing phase space structures of general time dependent dynamical sys- tems. Communications in Nonlinear Science and Numerical Simulation , 18(12):3530– 3557, 2013. doi:10.1016/j.cnsn...

    2013 doi

  34. [42]

    Mattheakis, D

    M. Mattheakis, D. Sondak, A. S. Dogra, and P. Protopapas. Hamiltonian neu- ral networks for solving equations of motion. Phys. Rev. E , 105:065305, Jun 2022. doi:10.1103/PhysRevE.105.065305

    2022 doi

  35. [43]

    Miguel Angel Bastarrachea-Magnani, L

    M. Miguel Angel Bastarrachea-Magnani, L. del Carpio B., J. Chávez-Carlos, S. Lerma- Hernández, and J. Hirsch. Regularity and chaos in cavity QED. Physica Scripta , 92(5):054003, 2017. doi:10.1088/1402-4896/aa6640

    2017 doi

  36. [44]

    H. T. Moges, T. Manos, O. Racoveanu, and C. Skokos. On the behavior of the gen- eralized alignment index (gali) method for dissipative systems, 2025. URL: https: //arxiv.org/abs/2503.01784, arXiv:2503.01784

    2025 arXiv

  37. [45]

    C. E. Montanari, R. B. Appleby, A. Bazzani, A. Fornara, M. Giovannozzi, S. Redaelli, G. Sterbini, and G. Turchetti. Chaos indicators for non-linear dynamics in circular particle accelerators. 4 2025.arXiv:2504.12741

    2025 arXiv

  38. [46]

    C. E. Montanari, A. Bazzani, M. Giovannozzi, and G. Turchetti. Using Dynamic Indicators for Probing Single-Particle Stability in Circular Accelerators. JACoW, IPAC2022:168–171, 2022. doi:10.18429/JACoW-IPAC2022-MOPOST042

    2022 doi

  39. [47]

    F. J. Muñoz Almaraz, E. Freire, and J. Galán-Vioque. Bifurcation behavior of the Furuta pendulum. International Journal of Bifurcation and Chaos , 17(08):2571–2578,

  40. [48]

    S. Naik, V. J. García-Garrido, and S. Wiggins. Finding nhim: Identifying high dimensional phase space structures in reaction dynamics using lagrangian descriptors. Communications in Nonlinear Science and Numerical Simulation , 79:104907, 2019. URL: https://www.sciencedirect.co...

    2019

  41. [49]

    Papaphilippou

    Y. Papaphilippou. Detecting chaos in particle accelerators through the frequency map analysis method. Chaos: An Interdisciplinary Journal of Nonlinear Science , 24(2):024412, 06 2014. arXiv:https://pubs.aip.org/aip/cha/article-pdf/doi/ 10.1063/1.4884495/19786722/024412\_1\_onl...

    2014 doi

  42. [50]

    L. A. Pocher, I. Haber, T. M. Antonsen, and P. G. O’Shea. Data-Driven Discovery of Beam Centroid Dynamics. 10 2024.arXiv:2410.14019. 9

    2024 arXiv

  43. [51]

    Prince and J

    P. Prince and J. Dormand. High order embedded runge-kutta formulae. Jour- nal of Computational and Applied Mathematics , 7(1):67–75, 1981. URL: https:// www.sciencedirect.com/science/article/pii/0771050X81900103, doi:10.1016/ 0771-050X(81)90010-3

    1981

  44. [52]

    I. I. Shevchenko. Dynamical Chaos in Planetary Systems , volume 463. 2020. doi: 10.1007/978-3-030-52144-8

    2020 doi

  45. [53]

    S. Sinha. Chaos and regularity in adaptive lattice dynamics.International Journal of Modern Physics B , 09(08):875–931, 1995. doi:10.1142/S0217979295000355

    1995 doi

  46. [54]

    C. Skokos. Alignment indices: a new, simple method for determining the ordered or chaotic nature of orbits.Journal of Physics A: Mathematical and General , 34(47):10029,

  47. [55]

    C. Skokos. The Lyapunov Characteristic Exponents and Their Computation , pages 63–135. Springer Berlin Heidelberg, Berlin, Heidelberg, 2010. doi:10.1007/ 978-3-642-04458-8_2

    2010

  48. [56]

    Skokos, C

    C. Skokos, C. Antonopoulos, T. Bountis, and M. Vrahatis. How does the Smaller Alignment Index (SALI) distinguish order from chaos?Progress of Theoretical Physics Supplement, 150:439–443, 2003. doi:10.1143/PTPS.150.439

    2003 doi

  49. [57]

    Skokos, C

    C. Skokos, C. Antonopoulos, T. Bountis, and M. Vrahatis. Detecting order and chaos in Hamiltonian systems by the SALI method.Journal of Physics A: Mathematical and General, 37(24):6269, 2004. doi:10.1088/0305-4470/37/24/006

    2004 doi

  50. [58]

    Skokos, T

    C. Skokos, T. Bountis, and C. Antonopoulos. Geometrical properties of local dynamics in Hamiltonian systems: The Generalized Alignment Index (GALI) method.Physica D: Nonlinear Phenomena , 231(1):30–54, 2007. doi:10.1016/j.physd.2007.04.004

    2007 doi

  51. [59]

    Skokos and E

    C. Skokos and E. Gerlach. Numerical integration of variational equations.Phys. Rev. E, 82:036704, 2010. doi:10.1103/PhysRevE.82.036704

    2010 doi

  52. [60]

    Skokos, G

    C. Skokos, G. Gottwald, and J. Laskar.Chaos Detection and Predictability . Lecture NotesinPhysics.SpringerBerlinHeidelberg, 2016. doi:10.1007/978-3-662-48410-4

    2016 doi

  53. [61]

    Skoufaris, J

    K. Skoufaris, J. Laskar, Y. Papaphilippou, and C. Skokos. Application of high order symplectic integration methods with forward integration steps in beam dynamics.Phys. Rev. Accel. Beams, 25:034001, Mar 2022. URL:https://link.aps.org/doi/10.1103/ PhysRevAccelBeams.25.034001, d...

    2022 doi

  54. [62]

    B. K. Tapley. Symplectic neural networks based on dynamical systems, 2024. URL: https://arxiv.org/abs/2408.09821, arXiv:2408.09821

    2024 arXiv

  55. [63]

    Y. Tong, S. Xiong, X. He, G. Pan, and B. Zhu. Symplectic neural networks in taylor series form for hamiltonian systems. Journal of Computational Physics , 437:110325, 2021. URL: https://www.sciencedirect.com/science/article/pii/ S0021999121002205, doi:10.1016/j.jcp.2021.110325

    2021

  56. [64]

    A. A. Tsonis and J. B. Elsner. Chaos, strange attractors, and weather. Bul- letin of the American Meteorological Society , 70(1):14 – 23, 1989. URL: https: //journals.ametsoc.org/view/journals/bams/70/1/1520-0477_1989_070_0014_ csaaw_2_0_co_2.xml, doi:10.1175/1520-0477(1989)07...

    1989 doi

  57. [65]

    Virtanen, R

    P. Virtanen, R. Gommers, T. E. Oliphant, M. Haberland, T. Reddy, D. Cournapeau, E. Burovski, P. Peterson, W. Weckesser, J. Bright, S. J. van der Walt, M. Brett, J. Wil- son, K. J. Millman, N. Mayorov, A. R. J. Nelson, E. Jones, R. Kern, E. Larson, C. J. Carey, İ. Polat, Y. Fen...

    2020

  58. [66]

    K. F. Wakker. Fundamentals of astrodynamics. 2015

    2015

  59. [67]

    A. Wolski. Beam Dynamics in High Energy Particle Accelerators . IMPERIAL COL- LEGE PRESS, 2014. URL: https://www.worldscientific.com/doi/abs/10.1142/ p899, arXiv:https://www.worldscientific.com/doi/pdf/10.1142/p899, doi:10. 1142/p899

    2014 doi

  60. [68]

    Zimper, A

    S. Zimper, A. Ngapasare, M. Hillebrand, M. Katsanikas, S. Wiggins, and S. Ch. Per- formance of chaos diagnostics based on lagrangian descriptors. application to the 4d standard map. Physica D: Nonlinear Phenomena , 453:133833, 2023. doi:10.1016/j. physd.2023.133833

    2023

  61. [69]

    E. E. Zotos. Distinguishing between order and chaos in a simple barred galaxy model. Astronomische Nachrichten, 338(5):614–620, 2017. doi:10.1002/asna.201713152. 11

    2017 doi

  62. [2001]

    doi:10.1088/0305-4470/34/47/309

    doi

  63. [2007]

    doi:10.1142/S0218127407018634

    doi

  64. [2016]

    doi:10.1021/acs.jpcb.5b09003

    doi

  65. [2020]

    URL: https://arxiv.org/abs/1909.13334, arXiv:1909.13334

    1909 arXiv

Pith tools

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