arxiv: 2604.19302 · v1 · submitted 2026-04-21 · ⚛️ physics.soc-ph · nlin.CD

Recognition: unknown

Node-weighted recurrence analysis for path dynamics on networks

A. Schmaus, N. Marwan, N. Molkenthin

Authors on Pith no claims yet

Pith reviewed 2026-05-10 01:25 UTC · model grok-4.3

classification ⚛️ physics.soc-ph nlin.CD

keywords recurrence analysisnetwork pathsbacktrackingrecurrence plotspath dynamicsnetwork structuremobilitynonlinear dynamics

0 comments

The pith

Recurrence plots of network paths display diagonal and perpendicular lines that mark backtracking and enable inferences about structure or dynamics.

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

The paper applies recurrence quantification analysis to trajectories of units moving on networks, such as polymers or human mobility paths. It shows that these paths generate recurrence patterns combining ordinary diagonal lines with perpendicular diagonal lines, a combination that points to backtracking along the same sequence of nodes. A sympathetic reader would care because the method offers a practical route to deduce either the underlying network topology from observed dynamics and plots or the dynamics from known networks and plots. This approach directly addresses the interaction between path behavior and network features in systems where full information is often incomplete.

Core claim

Trajectories of units moving on networks are relevant for nonlinear dynamical systems as diverse as polymers, ocean drifters, and human mobility. Although RQA is a well-researched tool with applications in many areas, it has rarely been used for spatial trajectories on networks. Here, we explore the use of RQA for paths on networks. We find that path dynamics on networks display recurrence patterns that are not often described in other applications of recurrence analysis. In particular, the combination of diagonal lines and perpendicular diagonal lines indicates backtracking paths. Recurrence analysis for path dynamics on networks can be helpful to better understand the network structure if

What carries the argument

Node-weighted recurrence plots constructed from path sequences, in which the joint presence of diagonal lines (repeated path segments) and perpendicular diagonal lines (backtracking on the same nodes) functions as the indicator of path dynamics on the network.

If this is right

When both the path dynamics and the recurrence plot are known, the underlying network structure can be inferred.
When both the network and the recurrence plot are known, the path dynamics can be inferred.
The method directly reveals the interaction between path dynamics and the topology of the network.

Where Pith is reading between the lines

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

The same recurrence signatures could be tested on empirical mobility traces to flag inefficient reversals without requiring the complete road network map in advance.
Extending the plots to weighted or time-resolved paths might distinguish between deliberate backtracking and random reversals in foraging or transport data.
Comparing recurrence patterns across random versus lattice or scale-free networks would quantify how topology modulates the visibility of backtracking.

Load-bearing premise

That the observed recurrence patterns of diagonal plus perpendicular lines are sufficiently distinctive and general to support reliable inferences about unknown network structure or dynamics from the plots alone, without case-specific validation.

What would settle it

A controlled path simulation on a known network that includes verified backtracking segments yet produces no perpendicular diagonal lines in the corresponding recurrence plot.

Figures

Figures reproduced from arXiv: 2604.19302 by A. Schmaus, N. Marwan, N. Molkenthin.

**Figure 2.** Figure 2: FIG. 2. Example trajectories on 7 [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** Figure 3: FIG. 3. Recurrence plots for all combinations of networks and path dynamics (for one path segment of 100 steps each). White areas indicate [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

**Figure 4.** Figure 4: FIG. 4. The five recurrence measures and the route lengths from [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

**Figure 5.** Figure 5: The NRR is plotted over node betweenness in Fig. 5 [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗

**Figure 5.** Figure 5: FIG. 5. Node-based recurrence measures vary with centrality. a) Node recurrence rate increases with node betweenness. Across all networks, [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗

read the original abstract

Trajectories of units moving on networks are relevant for nonlinear dynamical systems as diverse as polymers, ocean drifters, and human mobility. Although RQA is a well-researched tool with applications in many areas, it has rarely been used for spatial trajectories on networks. Here, we explore the use of RQA for paths on networks. We find that path dynamics on networks display recurrence patterns that are not often described in other applications of recurrence analysis. In particular, the combination of diagonal lines and perpendicular diagonal lines, indicates backtracking paths. We find that recurrence analysis for path dynamics on networks can be helpful to a) better understand the network structure if dynamic and recurrence plots are known, b) better understand the dynamics if network and recurrence plots are known, and c) understand the interaction between path dynamics and the underlying network.

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.

Desk Editor's Note private letter to a colleague

The paper adapts recurrence plots to network paths and flags a backtracking signature from combined diagonal and perpendicular lines, but the support for using those plots to infer unknown networks or dynamics is mostly observational.

read the letter

The core contribution is straightforward: they take recurrence quantification analysis, add node weighting to handle network structure, and apply it to trajectories that are forced to follow edges. The new observation is that backtracking produces a mix of diagonal lines and perpendicular diagonals in the recurrence plot, a combination they say is not common in other RQA settings. That pattern is worth noting for anyone who already works with recurrence plots on constrained paths like polymers or mobility traces. The three-way claim—that the plots can help decode the network when dynamics are known, decode the dynamics when the network is known, or illuminate their interaction—is a reasonable way to frame the potential use case. It stays grounded in the actual plots rather than overclaiming a new theorem or algorithm. The paper does a clean job of showing the idea on simple examples without burying it in extra machinery. The main limitation is that the distinctiveness of the pattern is asserted from exploration rather than tested against alternatives. There are no null models shown, no quantitative RQA measures like determinism or laminarity broken out by path type, and no checks on whether other walk rules or topologies can generate the same perpendicular lines. Without those, the inference steps remain plausible but unproven. A reader who already has network trajectory data might still find the visualization useful as a quick diagnostic, but anyone hoping to reverse-engineer structure from plots alone would need more controls before relying on it. This is the sort of incremental methods paper that fits a specialized journal or a methods section in an applied physics outlet. It is coherent on its own terms and shows honest engagement with the literature on RQA, so it deserves a serious referee rather than a desk rejection. The referees can ask for the missing validation runs and decide how much weight the backtracking claim can carry.

Referee Report

3 major / 1 minor

Summary. The manuscript explores the application of recurrence quantification analysis (RQA) to trajectories of units moving on networks, introducing a node-weighted variant of RQA. It reports that path dynamics on networks produce recurrence patterns not commonly seen in other RQA applications, specifically that the combination of diagonal lines and perpendicular diagonal lines in recurrence plots indicates backtracking paths. The work suggests that this approach can help (a) infer network structure from known dynamics and recurrence plots, (b) infer dynamics from known network and recurrence plots, and (c) understand the interaction between path dynamics and network structure.

Significance. If the reported recurrence patterns prove to be distinctive signatures of backtracking and the inferences can be made reliably, the work could extend RQA as a diagnostic tool for network-constrained dynamics in areas such as human mobility, polymer physics, and ocean drifters. The exploratory framing and introduction of node-weighted RQA represent a potential methodological contribution, but the absence of quantitative validation or controls limits the immediate significance.

major comments (3)

[Abstract] Abstract: The central claim that 'the combination of diagonal lines and perpendicular diagonal lines indicates backtracking paths' is presented without any supporting figures, quantitative RQA measures (e.g., determinism, laminarity, or recurrence rate), or explicit examples from specific networks or path types.
[Abstract] Abstract: The three listed applications (a, b, c) for inferring unknown network structure or dynamics from recurrence plots alone are asserted without demonstration of how the patterns enable such inferences, such as through controlled examples, null models (e.g., non-backtracking random walks on the same graphs), or algorithms for reverse-engineering structure from plots.
[Abstract] The manuscript does not define or derive the node-weighted recurrence analysis formally (e.g., via an equation for the recurrence matrix that incorporates node weights), leaving unclear how it differs from standard RQA and whether the perpendicular-line pattern is an artifact of the weighting or a general feature of network paths.

minor comments (1)

[Abstract] The abstract uses the term 'node-weighted recurrence analysis' without prior definition or reference to how node weights are assigned or incorporated into the recurrence plot construction.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their thoughtful and constructive comments on our manuscript. We address each of the major comments point by point below, providing clarifications and outlining the revisions we will implement.

read point-by-point responses

Referee: [Abstract] Abstract: The central claim that 'the combination of diagonal lines and perpendicular diagonal lines indicates backtracking paths' is presented without any supporting figures, quantitative RQA measures (e.g., determinism, laminarity, or recurrence rate), or explicit examples from specific networks or path types.

Authors: The abstract summarizes the key findings, but the full manuscript contains figures and detailed examples from specific networks and path types that demonstrate these recurrence patterns. Quantitative measures such as determinism and laminarity are computed and discussed in the results section. To better support the abstract claim, we will revise the abstract to include references to these quantitative measures and a brief mention of the examples used. revision: partial
Referee: [Abstract] Abstract: The three listed applications (a, b, c) for inferring unknown network structure or dynamics from recurrence plots alone are asserted without demonstration of how the patterns enable such inferences, such as through controlled examples, null models (e.g., non-backtracking random walks on the same graphs), or algorithms for reverse-engineering structure from plots.

Authors: The manuscript explores these applications through illustrative examples in the main text, showing how the observed patterns can aid in understanding the interplay. However, we agree that additional controlled experiments and null models would provide stronger support. We will add a new subsection with comparisons to non-backtracking walks and other controls to demonstrate the inferences more rigorously. revision: yes
Referee: [Abstract] The manuscript does not define or derive the node-weighted recurrence analysis formally (e.g., via an equation for the recurrence matrix that incorporates node weights), leaving unclear how it differs from standard RQA and whether the perpendicular-line pattern is an artifact of the weighting or a general feature of network paths.

Authors: We have defined the node-weighted recurrence matrix in the Methods section, where the standard recurrence matrix is modified by incorporating node weights to account for the network structure. The perpendicular line patterns are shown to arise from the backtracking dynamics on the network, independent of the weighting in our analysis. To address the concern, we will move or duplicate the formal definition to the introduction and add a derivation or explanation of how it differs from standard RQA. revision: yes

Circularity Check

0 steps flagged

No circularity: exploratory application of RQA to network paths is self-contained

full rationale

The paper is an observational exploration of recurrence quantification analysis applied to trajectories on networks. It reports empirical patterns (diagonal lines combined with perpendicular diagonals indicating backtracking) directly from the recurrence plots without any derivation chain that reduces to fitted inputs, self-definitions, or self-citation load-bearing premises. No equations, parameter fits, uniqueness theorems, or ansatzes are invoked that loop back to the inputs by construction. The three listed utilities (a-c) follow from the observed patterns as interpretive suggestions rather than forced predictions. This is the standard non-circular outcome for an exploratory methods paper.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

Abstract-only review yields no explicit free parameters, detailed axioms, or invented entities beyond the high-level framing of node-weighted RQA; the central claims rest on the domain assumption that recurrence plots remain interpretable when applied to network paths.

axioms (1)

domain assumption Recurrence plots constructed from network path trajectories can reveal interpretable structural and dynamic information
Invoked by the claim that the method helps understand network structure or dynamics from plots.

invented entities (1)

Node-weighted recurrence analysis no independent evidence
purpose: Adapt standard RQA to account for node properties when analyzing paths on networks
Presented as the core new technique in the title and abstract.

pith-pipeline@v0.9.0 · 5441 in / 1322 out tokens · 32151 ms · 2026-05-10T01:25:58.360881+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

30 extracted references · 1 canonical work pages · 1 internal anchor

[1]

comprehensive recurrence quantification analysis of path dynamics on networks
[2]

distinguishing between network and dynamics effects
[3]

Node-weighted recurrence analysis for path dynamics on networks

introduction of node-weighted tools for this analysis The trajectories are given as lists of nodes, which are visited in order. The trajectories are denoted asT={u 0,u 1...,u f }, whereu i is the numerical identifier of the node visited at time stepi. We denote the total length of the trajectory as|T|. To account for the spatial aspect of the trajectories...

work page internal anchor Pith review Pith/arXiv arXiv 2026
[4]

Flory theory for polymers.Journal of Physics: Con- densed Matter, 25(50):503101, 2013

Somendra M Bhattacharjee, Achille Giacometti, and Amos Maritan. Flory theory for polymers.Journal of Physics: Con- densed Matter, 25(50):503101, 2013

2013
[5]

Scaling laws in spatial net- work formation.Physical review letters, 117(16):168301, 2016

Nora Molkenthin and Marc Timme. Scaling laws in spatial net- work formation.Physical review letters, 117(16):168301, 2016

2016
[6]

The ontogeny of bumblebee flight trajectories: from naïve ex- plorers to experienced foragers.Plos one, 8(11):e78681, 2013

Juliet L Osborne, Alan Smith, Suzanne J Clark, Don R Reynolds, Mandy C Barron, Ka S Lim, and Andy M Reynolds. The ontogeny of bumblebee flight trajectories: from naïve ex- plorers to experienced foragers.Plos one, 8(11):e78681, 2013

2013
[7]

Anal- ysis of movement recursions to detect reproductive events and estimate their fate in central place foragers.Movement Ecology, 8(1):24, 2020

Simona Picardi, Brian J Smith, Matthew E Boone, Peter C Fred- erick, Jacopo G Cecere, Diego Rubolini, Lorenzo Serra, Si- mone Pirrello, Rena R Borkhataria, and Mathieu Basille. Anal- ysis of movement recursions to detect reproductive events and estimate their fate in central place foragers.Movement Ecology, 8(1):24, 2020

2020
[8]

Overview of ocean based buoys and drifters: present applications and future needs

NN Soreide, CE Woody, and SM Holt. Overview of ocean based buoys and drifters: present applications and future needs. InMTS/IEEE Oceans 2001. An Ocean Odyssey. Conference Proceedings (IEEE Cat. No. 01CH37295), volume 4, pages 2470–2472. IEEE, 2001

2001
[9]

Applying dynamical systems techniques to real ocean drifters.Nonlinear processes in geophysics, 29(4):345– 361, 2022

Irina I Rypina, Timothy Getscher, Lawrence J Pratt, and Tamay Ozgokmen. Applying dynamical systems techniques to real ocean drifters.Nonlinear processes in geophysics, 29(4):345– 361, 2022

2022
[10]

Traffic and related self-driven many-particle sys- 7 tems.Reviews of modern physics, 73(4):1067, 2001

Dirk Helbing. Traffic and related self-driven many-particle sys- 7 tems.Reviews of modern physics, 73(4):1067, 2001

2001
[11]

Scaling laws of collective ride-sharing dynamics.Physical Review Let- ters, 125(24):248302, 2020

Nora Molkenthin, Malte Schröder, and Marc Timme. Scaling laws of collective ride-sharing dynamics.Physical Review Let- ters, 125(24):248302, 2020

2020
[12]

Random walk models in biology.Journal of the Royal society interface, 5(25):813–834, 2008

Edward A Codling, Michael J Plank, and Simon Benhamou. Random walk models in biology.Journal of the Royal society interface, 5(25):813–834, 2008

2008
[13]

Protein contact maps: a binary depiction of protein 3d structures.Physica A: Statistical Mechanics and its Applications, 465:782–791, 2017

Isaac Arnold Emerson and Arumugam Amala. Protein contact maps: a binary depiction of protein 3d structures.Physica A: Statistical Mechanics and its Applications, 465:782–791, 2017

2017
[14]

Recurrence plots for the analysis of complex systems

Norbert Marwan, M Carmen Romano, Marco Thiel, and Jürgen Kurths. Recurrence plots for the analysis of complex systems. Physics reports, 438(5-6):237–329, 2007

2007
[15]

Marwan and K

N. Marwan and K. H. Kraemer. Trends in recurrence analysis of dynamical systems.European Physical Journal – Special Topics, 232:5–27, 2023

2023
[16]

C. L. Webber, Jr., A. Giuliani, J. P. Zbilut, and A. Colosimo. Elucidating protein secondary structures using alpha-carbon re- currence quantifications.Proteins: Structure, Function, and Genetics, 44(3):292–303, 2001

2001
[17]

W. I. Karain. Detecting transitions in protein dynamics using a recurrence quantification analysis based bootstrap method. BMC Bioinformatics, 18(1):525, 2017

2017
[18]

What geometrically constrained models can tell us about real-world protein contact maps.Physical Biology, 20(4):046004, 2023

J Jasmin Güven, Nora Molkenthin, Steffen Mühle, and Anto- nia SJS Mey. What geometrically constrained models can tell us about real-world protein contact maps.Physical Biology, 20(4):046004, 2023

2023
[19]

Lesne, J

A. Lesne, J. Mozziconacci, and S. Ramdani. Spatial Recon- struction from Contact Networks: A Unified Approach for Dy- namics and Genomics. In Y . Hirata, M. Shiro, M. Fukino, C. L. Webber Jr., K. Aihara, and N. Marwan, editors,Recur- rence Plots and Their Quantifications: Methodological Break- throughs and Interdisciplinary Discoveries, pages 131–145. Spr...

2025
[20]

Chromatic and anisotropic cross-recurrence quantification analysis of interper- sonal behavior

Ralf FA Cox, Steffie van der Steen, Marlenny Guevara, Lisette de Jonge-Hoekstra, and Marijn van Dijk. Chromatic and anisotropic cross-recurrence quantification analysis of interper- sonal behavior. InRecurrence Plots and Their Quantifica- tions: Expanding Horizons: Proceedings of the 6th Interna- tional Symposium on Recurrence Plots, Grenoble, France, 17-...

2015
[21]

Symbolic recurrence plots to analyze dynamical systems.Chaos: An Interdisciplinary Journal of Nonlinear Sci- ence, 28(6), 2018

M Caballero-Pintado, Mariano Matilla-García, and Manuel Ruiz Marín. Symbolic recurrence plots to analyze dynamical systems.Chaos: An Interdisciplinary Journal of Nonlinear Sci- ence, 28(6), 2018

2018
[22]

The structure and function of complex net- works.SIAM review, 45(2):167–256, 2003

Mark EJ Newman. The structure and function of complex net- works.SIAM review, 45(2):167–256, 2003

2003
[23]

Mixing of random walks and other diffusions on a graph.London Mathematical Society Lecture Note Series, pages 119–154, 1995

László Lovász and Peter Winkler. Mixing of random walks and other diffusions on a graph.London Mathematical Society Lecture Note Series, pages 119–154, 1995

1995
[24]

Random walks and flights over connected graphs and complex networks.Communications in Nonlinear Science and Numerical Simulation, 16(1):21–55, 2011

Dimitry V olchenkov. Random walks and flights over connected graphs and complex networks.Communications in Nonlinear Science and Numerical Simulation, 16(1):21–55, 2011

2011
[25]

An urban shared pooled mo- bility system cuts distance travelled by over 50%.Transporta- tion Research Part D: Transport and Environment, 144, 2025

Alexander Schmaus, Felix Creutzig, Nicolas Koch, Florian Nachtigall, and Nora Molkenthin. An urban shared pooled mo- bility system cuts distance travelled by over 50%.Transporta- tion Research Part D: Transport and Environment, 144, 2025

2025
[26]

A markov chain monte carlo approach for stop pooling in on-demand ride-pooling services

Alexander Schmaus. A markov chain monte carlo approach for stop pooling in on-demand ride-pooling services. 2022

2022
[27]

S. R. García. Chaos for examining the fundamental period of soils. pages 1341–1357, 2014

2014
[28]

Ramdani, F

S. Ramdani, F. Bouchara, J. Lagarde, and A. Lesne. Recurrence plots of discrete-time Gaussian stochastic processes.Physica A, 330:17–31, 2016

2016
[29]

Ramdani, F

S. Ramdani, F. Bouchara, and A. Lesne. Probabilistic analy- sis of recurrence plots generated by fractional Gaussian noise. Chaos, 28(8):085721, 2018

2018
[30]

Selec- tion of recurrence threshold for signal detection.The european physical journal special topics, 164(1):45–53, 2008

Stefan Schinkel, Olaf Dimigen, and Norbert Marwan. Selec- tion of recurrence threshold for signal detection.The european physical journal special topics, 164(1):45–53, 2008

2008