arxiv: 2604.04459 · v1 · submitted 2026-04-06 · ⚛️ physics.soc-ph · cs.CY

Recognition: 3 theorem links

· Lean Theorem

Intercity mobility reveals the hyperbolic geometry of city systems

Bin Liu, Changcheng Kan, Chenglong Wang, Kaixiang Zhang, Pengjun Zhao, Xiang Li, Xingjian Liu, Zhaoya Gong

Authors on Pith no claims yet

Pith reviewed 2026-05-10 19:41 UTC · model grok-4.3

classification ⚛️ physics.soc-ph cs.CY

keywords intercity mobilityhyperbolic geometrycity systemsurban hierarchyhinterland relationsspatial networksproximity and rank

0 comments

The pith

Intercity mobility flows embed into a latent hyperbolic geometry that captures the interplay between city hierarchy and proximity.

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

The paper develops a geometric model that places intercity mobility patterns inside a hyperbolic space, where distances and angles encode both the ranking of cities and their closeness without forcing a choice between the two. A sympathetic reader would care because this single structure explains why some cities draw wide but shallow hinterlands while others nest smaller ones more tightly, and why those patterns shift as one moves up the hierarchy. The authors test the model on twelve nationwide mobility datasets and report that hierarchies arise from the bottom up rather than being imposed from the top. They trace the non-stationary character of city-hinterland relations to ongoing trade-offs between rank and range. Hierarchy-dominated or proximity-dominated regimes can therefore be read off the changing balance of those trade-offs.

Core claim

Embedding intercity mobility into latent hyperbolic geometry unravels the measures of hierarchy and proximity while accounting for their interplay. Validation against twelve nationwide datasets shows a bottom-up emergence of city hierarchies; along this hierarchy the nesting and range properties of city-hinterland relations are non-stationary. These variations originate in trade-offs between city hierarchy and hinterland range, so that hierarchy- and proximity-dominated urban processes become distinguishable by examining the dynamics of the trade-offs.

What carries the argument

The latent hyperbolic geometry model that embeds observed mobility flows so that both rank order and spatial closeness are represented by the same curved metric.

If this is right

City hierarchies form bottom-up along the directions of strongest mobility rather than being preset by administrative rank.
City-hinterland relations change their nesting depth and geographic reach as one moves up the hierarchy.
Trade-offs between rank and hinterland range determine whether a given urban process is dominated by hierarchy or by proximity.
The same geometric coordinates allow direct comparison of hierarchy- versus proximity-dominated regimes across different national systems.

Where Pith is reading between the lines

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

The model supplies a concrete coordinate system that could be used to simulate how adding or removing transport links would shift the entire hierarchy and its hinterlands.
Because the geometry is low-dimensional, it offers a compact way to compare city-system evolution across countries or across time periods using only mobility matrices.
If the trade-off mechanism holds at smaller scales, the same embedding could be applied to intra-city flow data to detect neighborhood hierarchies.

Load-bearing premise

Real-world intercity mobility flows can be faithfully represented in a latent hyperbolic space whose geometry directly encodes the interplay between hierarchy and proximity without substantial distortion.

What would settle it

Apply the embedding procedure to a fresh nationwide mobility dataset; if the resulting hyperbolic coordinates fail to recover observed city ranks, hinterland boundaries, or the reported non-stationarity of nesting and range, the claim is falsified.

read the original abstract

The hierarchy and proximity are key dimensions of urban relational processes, but their interplay in shaping intercity interactions and the underlying structures of city systems remain unclear. We develop a novel geometric model of city systems embedding intercity mobility into a latent hyperbolic geometry, which unravels the measures of hierarchy and proximity accounting for their interplay. It is successfully validated against 12 different nationwide intercity mobility datasets. We find a bottom-up emergence of city hierarchies, along which the variations of city-hinterland relations are non-stationary in terms of their nesting and range properties. Such non-stationarity originates from trade-offs between city hierarchy and hinterland range in determining the formation of city-hinterland structures. Hierarchy- and proximity-dominated urban processes can be elucidated from examining dynamics of the trade-offs. The revealed urban relational processes of city systems are at the core of the emerging science of cities and crucial for spatial planning and regional policymaking.

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

Hyperbolic embedding of intercity mobility produces plausible hierarchies on 12 datasets, but the paper does not show why negative curvature is necessary rather than optional.

read the letter

The core contribution is an application of hyperbolic embedding to intercity mobility flows. Radial coordinates are interpreted as city hierarchy and angular separation as proximity, with the model then used to describe how city-hinterland nesting changes along the hierarchy. The authors report consistent patterns across twelve national mobility datasets and note that the trade-off between hierarchy strength and hinterland range is not stationary. That multi-dataset scope is the clearest strength; it moves the claim beyond a single-country case study. The non-stationary finding is also worth attention if the embedding procedure does not force it by construction. The work sits in the existing literature on hyperbolic network embeddings but narrows the focus to mobility-derived city systems and extracts a specific dynamic about hinterland relations. Credit is due for the empirical breadth. The main limitation is the absence of direct comparisons. If Euclidean or mixed-curvature embeddings recover comparable hierarchy rankings and reconstruction accuracy on the same flows, the geometric interpretation loses uniqueness. The abstract and available description give no reconstruction-error tables, no baseline gravity-model fits, and no sensitivity checks on curvature or dimension. Without those, it is difficult to separate genuine geometric necessity from a convenient re-description of the input data. The fitting step itself is latent and therefore carries the usual risk that observed structure partly reflects modeling choices rather than independent signal. This paper is aimed at researchers who already work with latent-space models of spatial networks or with quantitative urban geography. A reader looking for new organizing principles in city systems will find the empirical patterns useful to discuss, even if the geometric framing needs more defense. It is worth sending to peer review because the dataset coverage is solid and the topic is relevant, but referees should be asked to examine the embedding validation and alternative-model tests in detail.

Referee Report

3 major / 2 minor

Summary. The paper develops a geometric model that embeds intercity mobility flows into a latent hyperbolic space, with radial coordinates encoding city hierarchy and angular separations encoding proximity, thereby accounting for their interplay in shaping city systems. The model is validated on 12 nationwide intercity mobility datasets and yields findings of bottom-up emergence of city hierarchies together with non-stationary variations in city-hinterland nesting and range that arise from explicit trade-offs between hierarchy and hinterland extent.

Significance. If the embedding is shown to be faithful, low-distortion, and uniquely required by the data, the work would supply a concrete geometric mechanism linking mobility to hierarchical urban structure, with direct relevance to the science of cities and to spatial planning. The multi-dataset validation and the focus on non-stationary trade-offs are potentially valuable strengths, provided they survive quantitative scrutiny against alternative geometries.

major comments (3)

[Validation] Validation section (and any associated tables/figures reporting fit quality): the claim of successful validation on 12 datasets must be accompanied by explicit quantitative metrics—e.g., embedding distortion, flow-reconstruction error, or log-likelihood—together with direct comparisons to Euclidean and mixed-curvature embeddings on the identical datasets. Without these, it remains unclear whether hyperbolic geometry is necessary or merely sufficient for capturing the reported hierarchy-proximity trade-offs.
[Methods] Methods section describing the embedding procedure: the paper should demonstrate that the inferred radial (hierarchy) coordinates and the detected non-stationarity are robust to reasonable variations in embedding dimension, curvature, and optimization hyperparameters. If the geometry is obtained by maximum-likelihood fitting to the same mobility volumes used to define hierarchy, an explicit check against circularity is required.
[Results] Results section on city-hinterland relations: the assertion that non-stationarity originates from hierarchy-hinterland-range trade-offs needs supporting statistical evidence (e.g., regression coefficients, permutation tests, or time-series stationarity diagnostics) showing that the observed nesting and range variations are driven by the geometric trade-off rather than by dataset-specific confounders or post-hoc partitioning choices.

minor comments (2)

[Abstract] Abstract: the phrasing 'unravels the measures of hierarchy and proximity' is vague; a single sentence clarifying the operational definitions of these two quantities would improve readability.
[Figures] Figure captions and legends: ensure that all panels reporting embedding coordinates or trade-off surfaces include axis labels, curvature values, and sample sizes so that readers can assess the scale of the reported effects.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the detailed and constructive report. The comments identify important gaps in quantitative validation, robustness, and statistical support that we agree require attention. We outline targeted revisions below to address each major point directly while preserving the core contributions of the hyperbolic embedding framework.

read point-by-point responses

Referee: [Validation] Validation section (and any associated tables/figures reporting fit quality): the claim of successful validation on 12 datasets must be accompanied by explicit quantitative metrics—e.g., embedding distortion, flow-reconstruction error, or log-likelihood—together with direct comparisons to Euclidean and mixed-curvature embeddings on the identical datasets. Without these, it remains unclear whether hyperbolic geometry is necessary or merely sufficient for capturing the reported hierarchy-proximity trade-offs.

Authors: We agree that the current validation claims would be substantially strengthened by explicit quantitative metrics and comparative benchmarks. In the revised manuscript we will expand the Validation section to report embedding distortion (e.g., average relative error and stress), flow-reconstruction error (mean absolute percentage error on held-out flows), and log-likelihood values for the hyperbolic model. We will also add side-by-side comparisons against Euclidean and mixed-curvature embeddings fitted to the identical 12 datasets, using the same optimization protocol. These additions will allow readers to assess whether hyperbolic geometry is required rather than merely sufficient. revision: yes
Referee: [Methods] Methods section describing the embedding procedure: the paper should demonstrate that the inferred radial (hierarchy) coordinates and the detected non-stationarity are robust to reasonable variations in embedding dimension, curvature, and optimization hyperparameters. If the geometry is obtained by maximum-likelihood fitting to the same mobility volumes used to define hierarchy, an explicit check against circularity is required.

Authors: We accept the need for explicit robustness checks. The revised Methods section will include a dedicated sensitivity analysis varying embedding dimension (2D baseline plus higher-dimensional trials), curvature values around the fitted optimum, and key hyperparameters (learning rate, batch size, convergence tolerance). To address circularity, we will add a comparison of the inferred radial coordinates against independent, externally sourced hierarchy proxies (city population rank and GDP) and will report results from a cross-validation scheme in which a subset of mobility flows is withheld during embedding. These steps will demonstrate that the reported non-stationarity is not an artifact of the fitting procedure. revision: yes
Referee: [Results] Results section on city-hinterland relations: the assertion that non-stationarity originates from hierarchy-hinterland-range trade-offs needs supporting statistical evidence (e.g., regression coefficients, permutation tests, or time-series stationarity diagnostics) showing that the observed nesting and range variations are driven by the geometric trade-off rather than by dataset-specific confounders or post-hoc partitioning choices.

Authors: We agree that the causal link between the geometric trade-off and the observed non-stationarity requires stronger statistical grounding. In the revised Results section we will introduce (i) multivariate regressions of hinterland nesting and range on radial coordinate (hierarchy) while controlling for dataset identity, (ii) permutation tests that randomly reassign cities to hierarchy bins and recompute nesting statistics, and (iii) Augmented Dickey-Fuller stationarity diagnostics on the time series of range and nesting measures. These analyses will quantify the contribution of the hierarchy-hinterland trade-off and help rule out confounding factors or arbitrary partitioning effects. revision: yes

Circularity Check

0 steps flagged

No circularity detected; derivation self-contained against data validation

full rationale

The abstract and available text describe embedding mobility flows into hyperbolic geometry, followed by validation on 12 independent nationwide datasets and analysis of resulting hierarchies and trade-offs. No equations, self-citations, or derivation steps are quoted that reduce the central geometric claims or predictions to fitted parameters by construction. The model is presented as novel but externally checked via reconstruction on held-out or separate data sources, satisfying the criteria for non-circularity. No self-definitional, fitted-input-renamed-as-prediction, or load-bearing self-citation patterns are exhibited.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The central model presupposes that mobility data admits a faithful hyperbolic embedding whose curvature encodes hierarchy-proximity trade-offs, but the precise assumptions and any fitted quantities remain unspecified.

pith-pipeline@v0.9.0 · 5473 in / 1237 out tokens · 51194 ms · 2026-05-10T19:41:48.300967+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking (D=3 from linking) unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We propose a geometric model of city systems by embedding intercity networks into the Poincaré ball space... radial and angular dimensions encode hierarchy and proximity, respectively. The Poincaré distance... d(u,v)≈r_u + r_v + 2 ln(sin(Δθ/2))
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel (J uniquely satisfies Aczél functional equation) unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

minimize the loss function L = -∑ log( e^{-d(u,v)} / (e^{-d(u,v)} + ∑ e^{-d(u,v')}) ) + λ max r_v
IndisputableMonolith/Foundation/ArithmeticFromLogic.lean embed_eq_pow (LogicNat orbit under generator yields powers) unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

distributions of hierarchy measures... exhibit power law behavior... double pareto distributions

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

65 extracted references

[1]

Notably, all datasets cover the years before and after the outbreak of Covid-19

e., daily-travel (short-term) and annual-migration (long-term), from China and United States for years of 2019~2021: D1 and D3 are the daily-travel datasets for China and the US, respectively; D2 and D4 are the annual-migration datasets for China (years of 2019~2023) and the US, respectively (see ‘Data and pre-processing’ in SI Methods). Notably, all data...

2019
[2]

Despite that, we did not find a consistent relationship between the variation of range and that of NF by city order across all datasets (Fig

the larger deviation of concentration of residuals from zero, which implies that City-hinterland structures in China exhibit more non-stationarity in terms of hinterland range than those in the US. Despite that, we did not find a consistent relationship between the variation of range and that of NF by city order across all datasets (Fig. S6), as the range...

2019
[3]

For long-term datasets of China, the hierarchy-dominated urban process was weakened by the impact of Covid-19 related non-pharmaceutical interventions not only on travel but also on economic activities.39,40 Consistently, Chongqing tree has a stronger hierarchy-dominated urban process than Hangzhou tree does even though both trees have structural changes ...

2020
[4]

Distribution of NF for short-term and long-term datasets of China (a) and the US (e)

Nesting factor (NF) and range of city-hinterland structures in city systems. Distribution of NF for short-term and long-term datasets of China (a) and the US (e). Red (blue) dot and line represent the median (mean) of NF to highlight its variation over time. Mean and standard deviation (std) of NF by city order for short-term and long-term datasets of Chi...

2019
[5]

Theories of systems of cities

H. M. Abdel-Rahman, A. Anas, “Theories of systems of cities” in Handbook of Regional and Urban Economics, (Elsevier, 2004), pp. 2293–2339

2004
[6]

G. F. Mulligan, M. D. Partridge, J. I. Carruthers, Central place theory and its reemergence in regional science. Ann Reg Sci 48, 405–431 (2012)

2012
[7]

P. J. Taylor, M. Hoyler, R. Verbruggen, External urban relational process: Introducing central flow theory to complement central place theory. Urban Stud. 47, 2803–2818 (2010)

2010
[8]

Batty, The new science of cities (MIT press, 2013)

M. Batty, The new science of cities (MIT press, 2013)

2013
[9]

Batty, Rank clocks

M. Batty, Rank clocks. Nat 444, 592–596 (2006)

2006
[10]

L. M. Bettencourt, The origins of scaling in cities. Sci 340, 1438–1441 (2013)

2013
[11]

Bassolas, et al., Hierarchical organization of urban mobility and its connection with city livability

A. Bassolas, et al., Hierarchical organization of urban mobility and its connection with city livability. Nat. Commun. 10, 4817 (2019)

2019
[12]

Xu, et al., Urban dynamics through the lens of human mobility

Y. Xu, et al., Urban dynamics through the lens of human mobility. Nat. Comput. Sci. 3, 611–620 (2023)

2023
[13]

Capello, The city network paradigm: measuring urban network externalities

R. Capello, The city network paradigm: measuring urban network externalities. Urban Stud. 37, 1925–1945 (2000)

1925
[14]

Derudder, P

B. Derudder, P. J. Taylor, Central flow theory: Comparative connectivities in the world-city network. Reg. Stud. 52, 1029–1040 (2018)

2018
[15]

P. J. Taylor, Specification of the world city network. Geogr. Anal. 33, 181–194 (2001)

2001
[16]

García-Pérez, M

G. García-Pérez, M. Boguñá, A. Allard, M. Á. Serrano, The hidden hyperbolic geometry of international trade: World Trade Atlas 1870–2013. Sci Rep 6, 33441 (2016)

2013
[17]

Arcaute, et al., Cities and regions in Britain through hierarchical percolation

E. Arcaute, et al., Cities and regions in Britain through hierarchical percolation. R. Soc. Open Sci. 3, 150691 (2016)

2016
[18]

Brockmann, D

D. Brockmann, D. Helbing, The hidden geometry of complex, network-driven contagion phenomena. Sci 342, 1337–1342 (2013). 18

2013
[19]

Beguin, Christaller’s central place postulates: a commentary

H. Beguin, Christaller’s central place postulates: a commentary. Ann Reg Sci 26, 209–229 (1992)

1992
[20]

Tabuchi, J.-F

T. Tabuchi, J.-F. Thisse, A new economic geography model of central places. J. Urban Econ. 69, 240–252 (2011)

2011
[21]

Gabaix, Zipf’s law for cities: an explanation

X. Gabaix, Zipf’s law for cities: an explanation. Q J Econ 114, 739–767 (1999)

1999
[22]

The evolution of city size distributions

X. Gabaix, Y. M. Ioannides, “The evolution of city size distributions” in Handbook of Regional and Urban Economics, (Elsevier, 2004), pp. 2341–2378

2004
[23]

Duranton, Some foundations for Zipf’s law: Product proliferation and local spillovers

G. Duranton, Some foundations for Zipf’s law: Product proliferation and local spillovers. Reg. Sci. Urban Econ. 36, 542–563 (2006)

2006
[24]

Rossi-Hansberg, M

E. Rossi-Hansberg, M. L. Wright, Urban structure and growth. Rev Econ Stud 74, 597–624 (2007)

2007
[25]

Batty, The size, scale, and shape of cities

M. Batty, The size, scale, and shape of cities. Sci 319, 769–771 (2008)

2008
[26]

H. A. Makse, S. Havlin, H. E. Stanley, Modelling urban growth patterns. Nat 377, 608–612 (1995)

1995
[27]

Castells, The rise of the network society (John wiley & sons, 2011)

M. Castells, The rise of the network society (John wiley & sons, 2011)

2011
[28]

Alternative explanations of hierarchical differentiation in urban systems

D. Pumain, “Alternative explanations of hierarchical differentiation in urban systems” in Hierarchy in Natural and Social Sciences, (Springer, 2006), pp. 169–222

2006
[29]

E. J. Meijers, M. J. Burger, M. M. Hoogerbrugge, Borrowing size in networks of cities: City size, network connectivity and metropolitan functions in Europe. Papers in regional science 95, 181–199 (2016)

2016
[30]

Louail, et al., Uncovering the spatial structure of mobility networks

T. Louail, et al., Uncovering the spatial structure of mobility networks. Nat. Commun. 6, 6007 (2015)

2015
[31]

Pappalardo, E

L. Pappalardo, E. Manley, V. Sekara, L. Alessandretti, Future directions in human mobility science. Nat. Comput. Sci. 3, 588–600 (2023)

2023
[32]

S. M. Reia, P. S. C. Rao, M. Barthelemy, S. V. Ukkusuri, Domestic migration and city rank dynamics. Nat. Cities 2, 38–46 (2025)

2025
[33]

Rushton, Analysis of Spatial Behavior by Revealed Space Preference

G. Rushton, Analysis of Spatial Behavior by Revealed Space Preference. Ann. Assoc. Am. Geogr. 59, 391–400 (1969)

1969
[34]

Boguna, et al., Network geometry

M. Boguna, et al., Network geometry. Nat. Rev. Phys. 3, 114–135 (2021)

2021
[35]

Papadopoulos, M

F. Papadopoulos, M. Kitsak, M. Á. Serrano, M. Boguná, D. Krioukov, Popularity versus similarity in growing networks. Nat 489, 537–540 (2012)

2012
[36]

Boguná, F

M. Boguná, F. Papadopoulos, D. Krioukov, Sustaining the internet with hyperbolic mapping. Nat. Commun. 1, 62 (2010)

2010
[37]

Krioukov, F

D. Krioukov, F. Papadopoulos, M. Kitsak, A. Vahdat, M. Boguná, Hyperbolic geometry of complex networks. Phys. Rev. E 82, 036106 (2010)

2010
[38]

Hyperbolic groups

M. Gromov, “Hyperbolic groups” in Essays in Group Theory, (Springer, 1987), pp. 75–263

1987
[39]

Nickel, D

M. Nickel, D. Kiela, Poincaré embeddings for learning hierarchical representations. Adv Neural Inf Process Syst 30 (2017)

2017
[40]

Krioukov, F

D. Krioukov, F. Papadopoulos, A. Vahdat, M. Boguná, Curvature and temperature of complex networks. Phys. Rev. E 80, 035101 (2009)

2009
[41]

Radford, et al., Learning transferable visual models from natural language supervision in International Conference on Machine Learning, (PmLR, 2021), pp

A. Radford, et al., Learning transferable visual models from natural language supervision in International Conference on Machine Learning, (PmLR, 2021), pp. 19 8748–8763

2021
[42]

Mitzenmacher, A brief history of generative models for power law and lognormal distributions

M. Mitzenmacher, A brief history of generative models for power law and lognormal distributions. Internet Math. 1, 226–251 (2004)

2004
[43]

Pangallo, et al., The unequal effects of the health–economy trade-off during the COVID-19 pandemic

M. Pangallo, et al., The unequal effects of the health–economy trade-off during the COVID-19 pandemic. Nat. Hum. Behav. 8, 264–275 (2024)

2024
[44]

D. J. Haw, et al., Optimizing social and economic activity while containing SARS-CoV-2 transmission using DAEDALUS. Nat. Comput. Sci. 2, 223–233 (2022)

2022
[45]

Batty, P

M. Batty, P. A. Longley, Fractal cities: a geometry of form and function (Academic press, 1994)

1994
[46]

Hsu, Central place theory and city size distribution

W.-T. Hsu, Central place theory and city size distribution. Econ. J. 122, 903–932 (2012)

2012
[47]

Louail, M

T. Louail, M. Barthelemy, A dominance tree approach to systems of cities. Comput. Environ. Urban Syst. 97, 101856 (2022)

2022
[48]

Alessandretti, U

L. Alessandretti, U. Aslak, S. Lehmann, The scales of human mobility. Nat 587, 402–407 (2020)

2020
[49]

L. M. Bettencourt, J. Lobo, D. Helbing, C. Kühnert, G. B. West, Growth, innovation, scaling, and the pace of life in cities. Proc. Natl. Acad. Sci. 104, 7301–7306 (2007)

2007
[50]

Li, et al., Simple spatial scaling rules behind complex cities

R. Li, et al., Simple spatial scaling rules behind complex cities. Nat. Commun. 8, 1841 (2017)

2017
[51]

Sarkar, et al., Evidence for localization and urbanization economies in urban scaling

S. Sarkar, et al., Evidence for localization and urbanization economies in urban scaling. R. Soc. Open Sci. 7, 191638 (2020)

2020
[52]

Pumain, F

D. Pumain, F. Paulus, C. Vacchiani-Marcuzzo, J. Lobo, An evolutionary theory for interpreting urban scaling laws. Cybergeo: Eur. J. Geogr. (2006)

2006
[53]

Rushton, R

G. Rushton, R. G. Golledge, W. A. Clark, Formulation and test of a normative model for the spatial allocation of grocery expenditures by a dispersed population. Ann. Assoc. Am. Geogr. 57, 389–400 (1967)

1967
[54]

Rushton, Behavioral correlates of urban spatial structure

G. Rushton, Behavioral correlates of urban spatial structure. Econ. Geogr. 47, 49–58 (1971)

1971
[55]

Kang, Y. et al. Multiscale dynamic human mobility flow dataset in the US during the COVID-19 epidemic. Sci. Data. 7, 390 (2020)

2020
[56]

Stochastic gradient descent on Riemannian manifolds

Bonnabel, S. Stochastic gradient descent on Riemannian manifolds. IEEE Transactions on Automatic Control 58, 2217–2229 (2013)

2013
[57]

B., Sullivan, B

Adcock, A. B., Sullivan, B. D. & Mahoney, M. W. Tree-like structure in large social and information networks. in 2013 IEEE 13th international conference on data mining 1–10 (IEEE, 2013)

2013
[58]

Die Zentralen Orte in Süddeutschland: Eine Ökonomisch-Geogr

Christaller, W. Die Zentralen Orte in Süddeutschland: Eine Ökonomisch-Geogr. Unters. Über d. Gesetzmässigkeit d. Verbreitg u. Entwicklg d. Siedlgn Mit Städt. Funktionen. (Fischer, 1933)

1933
[59]

A., Hancock, J

Kiskowski, M. A., Hancock, J. F. & Kenworthy, A. K. On the use of Ripley’s K-function and its derivatives to analyze domain size. Biophys. J. 97, 1095–1103 (2009). 20 Supporting Information for Intercity mobility reveals the hyperbolic geometry of city systems Zhaoya Gong1,2, Bin Liu1,2,†, Chenglong Wang1,2,†, Pengjun Zhao1,2,*, Xiang Li1,2, Kaixiang Zhan...

2009
[60]

The real-time location information of users is derived from the geographic position associated with their requests

The data includes details such as origin, destination, date, and the number of movements. The real-time location information of users is derived from the geographic position associated with their requests. Baidu has enhanced the spatial accuracy of this data through data cleaning and algorithmic optimization. In terms of user sample representation, the da...

2019
[61]

A migration event is identified when a user resides in a new city for an extended period (typically several months)

It includes details such as origin city, destination city, and number of migrants, based on geographic positions from user requests. A migration event is identified when a user resides in a new city for an extended period (typically several months). This dataset is used directly to analyze long-term migration patterns. The short-term mobility dataset for ...

2019
[62]

For analysis, daily movements over the 31-day period are averaged to estimate mean daily mobility

The dataset includes details such as origin county, destination county, date, and number of movements, derived from anonymized mobile device location data. For analysis, daily movements over the 31-day period are averaged to estimate mean daily mobility. County-level flows are aggregated to the MSA level. The long-term mobility dataset for the US (D4) is ...

2020
[63]

can be expressed as: 𝛬(𝑟̅

We compute the discrete slope of 𝑑±(𝑟¯-) between adjacent orders and the Change Rate of Interaction Threshold (CRIT) is defined as the mean of these adjacent-order slopes: CRIT=1𝐼º𝑑±(𝑟¯-.,)−𝑑±(𝑟¯-)𝑟¯-.,−𝑟¯- N -B, (33) 32 Supporting Text Supplementary Note 1: Model comparison To assess the effectiveness of our proposed hyperbolic geometric model (HGM), we ...

2019
[64]

Higher values indicate better structural preservation. CN PSO HGM 20 30 40 50 20 30 40 50 Short-term 2019 0.51 0.57 0.59 0.61 0.67 0.67 0.66 0.67 2020 0.54 0.58 0.59 0.63 0.66 0.65 0.65 0.65 2021 0.52 0.56 0.59 0.61 0.62 0.62 0.62 0.63 Long-term 2019 0.39 0.46 0.50 0.54 0.49 0.50 0.58 0.59 2020 0.41 0.48 0.51 0.55 0.47 0.50 0.54 0.56 2021 0.46 0.50 0.55 0...

2019
[65]

Higher values indicate better structural preservation. US PSO HGM 20 30 40 50 20 30 40 50 Short-term 2019 0.36 0.44 0.49 0.54 0.53 0.56 0.59 0.62 2020 0.34 0.40 0.47 0.52 0.53 0.57 0.61 0.63 2021 0.37 0.44 0.49 0.53 0.50 0.53 0.56 0.60 Long-term 2019 0.36 0.43 0.48 0.51 0.27 0.31 0.31 0.30 2020 0.38 0.43 0.48 0.51 0.26 0.30 0.31 0.30 2021 0.38 0.42 0.48 0...

2019