Polygonal network disorder and the turning distance
Alexander Dolce, Ryan Lavelle, Bernard Scott, Ashlyn Urbanski, Joseph Klobusicky · 2025
Read Polygonal network disorder and the turning distance by Alexander Dolce, Ryan Lavelle, Bernard Scott, Ashlyn Urbanski, Joseph Klobusicky (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-06415.txt.
Introduction
Alex Dolce and four co-authors coin a score for how jumbled a web of polygons is, built on a known shape match tool. That tool, turning distance, weighs how alike two shapes are by the turns of their edges. Their 2025 study takes the mean gap between each face of a web and a tidy shape, like a regular polygon or a circle.
The team derives exact formulas for tidy shape pairs, sorted by how their side counts divide. Those formulas cut the math time for such pairs a great deal. Neat tilings get exact jumble scores. Tests on spring webs that reshuffle and on webs that tear show the score catches distinct kinds of disorder. Word choice of tidy shape and area weights shifts what the score sees.
Catalog record
| CREATOR | Alexander Dolce, Ryan Lavelle, Bernard Scott, Ashlyn Urbanski, Joseph Klobusicky |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-06415
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Polygonal network disorder and the turning distance Alexander Dolce, Ryan Lavelle, Bernard Scott, Ashlyn Urbanski, Joseph Klobusicky Abstract The turning distance is a well-studied metric for measuring the similarity between two polygons. This metric is constructed by taking an L p L^{p} distance between step functions which track each shape’s tangent angle of a path tracing its boundary. In this study, we introduce turning disorders for polygonal planar networks, defined by averaging turning distances between network faces and “ordered” shapes (regular polygons or circles). We derive closed-form expressions of turning distances for special classes of regular polygons, related to the divisibility of m m and n n , and also between regular polygons and circles. These formulas are used to show that the time for computing the 2-turning distances reduces to O ( ( m + n ) log ( m + n ) ) O((m+n)\log(m+n)) when both shapes are regular polygons, an improvement from O ( m n log ( m n ) ) O(mn\log(mn)) operations needed to compute distances between general polygons of n n and m m sides. We also apply these formulas to several examples of network microstructure with varying disorder. For Archimedean lattices, a class of regular tilings, we can express turning disorders with exact expressions. We also consider turning disorders applied to two examples of stochastic processes on networks: spring networks evolving under T1 moves and polygonal rupture processes. We find that the two aspects of defining different turning disorders, the choice of ordered shape and whether to apply area-weighting, can capture different notions of network disorder. keywords shape recognition, turning distance, microstructure, T1 move 1 Introduction In the field of computer vision, a popular topic of research focuses on how to measure the similarity between two shapes. For example, measures have been derived for comparing shapes to commonly found structures such as cubes Martinez-Ortiz and Žunić (2009) , rectangles Rosin (1999) , and circles Reeve (1979) . In two dimensions, Arkin et al. Arkin et al. (1989) provided a general metric, the turning distance , for comparing shapes by using L^{p} distances between turning functions which track tangent vectors along paths tracing each shape’s boundary. In polygons, turning functions are step functions, and it was found that turning distance under the L^{2} norm can be computed with a run time of O(mn\log(mn)) for comparing two polygons with m and n sides. In this paper, we will employ the turning function to define a notion of disorder in planar polygonal networks. We will frequently refer to such networks as microstructure, since approximately polygonal networks often appear in the study of grain boundary coarsening of polycrystalline metals in materials science Henseler et al. (2003) ; Epshteyn et al. (2022) ; Elsey and Slepčev (2015) . However, such patterns in nature are found on a variety of scales, ranging from pottery crazing Rice (2015) to Martian terrains Seibert and Kargel (2001) . While the notion of “disorder” can have multiple interpretations, here we will consider disorder of a network as an average of turning distances, taken between cells of the network and some predetermined “ordered shape”. For a polygon P with n sides, the three choices we will use for an ordered shape are the circle C , the regular hexagon R_{6} , and a regular polygon R_{n} with the same number of sides as P . For each of these measures, we also consider disorders with area-weighted averaging. While the various types of disorder appear to differ minimally in their definitions, we nonetheless find that these measures can vary widely given different examples of microstructures. In general, however, all of the disorders assign higher disorder to microstructures with nonconvex and irregular cells. In Sec. 2 , we review the turning distance, and subsequently introduce the various turning disorders, along with some of their basic properties. We delay an analysis of examples of the turning disorder, and instead focus first on formulas of turning distances where one of the shapes is either a regular polygon or a circle. Sec. 3 gives computationally efficient formulas for computing the turning distance. In the case of comparing two regular polygons with n and m sides, the turning distance formula can be simplified to a summation formula needing only O((m+n)\log(m+n)) operations. For Sec. 4 , further simplifications can be made when considering the divisibility between the number of sides of the regular polygons. In several cases, the turning distance can be written in closed form. A simple summation formula is also provided for comparing any polygon to a circle. This also reduces to closed form when the circle is compared to a regular polygon. The formulas derived in Secs. 2 and 3 can be used for efficiently computing several examples of microstructure. In Sec. 5 , we consider an ordered example of Archimedean lattices, a class of microstructures comprised of tiles which are regular polygons. Archimedean lattices arise in nanoscale science, including problems in self-assembly Kulkarni et al. (2020) and quantum magnets Farnell et al. (2014) . For soap foams with millimeter scale cells, Bae et al. Bae et al. (2019) gave methods for generating two-dimensional foams with Archimedean lattice structure. We write explicit expressions for each of the turning distance for the (4,8^{2}),(4,6,12), and (3,12^{2}) lattices. Sec. 6 studies two classes of stochastic processes on networks that produce increasingly disordered structures. In Sec. 6.1 , we consider linear spring networks in force balance. To evolve the network, we apply repeated applications of the T1 move, a simple topological transition which changes the connectivity of a network edge. A purely topological treatment of network statistics for the random T1 process was undertaken in the theoretical physics community Aste and Sherrington (1999) ; Sherrington et al. (2002) as a generalized version of the Ising model on planar graphs, and was found to exhibit glassy behavior under Metropolis-Hastings rules for rejecting certain configurations. By allowing the spring network to equilibrate after each T1 transition, we are now able to study the geometric properties of the spring network, and for our purposes we are able to compute turning disorders. In Sec. 6.2 , we randomly remove edges in a hexagonal lattice in a process topologically similar to edge rupturing in quasi-two-dimensional soap foams Klobusicky et al. (2024) . In this process, vertices are fixed and do not undergo equilibration after ruptures. This produces irregular, nonconvex shapes which result in large turning disorders. In Sec. 7 , we discuss potential areas in which turning measures can be applied to the problem of classification of materials. 2 A measure of disorder for microstructures A popular method for comparing shapes in two dimensions, proposed by Arkin et. al. Arkin et al. (1989) , tracks the tangent vector of a shape as it traces its boundary. A shape A can be scaled so that its perimeter is unit length, with its boundary traversed by a continuous curve \varphi(s):[0,1]\rightarrow\mathbb{R}^{2} parametrized by arc length. The turning function \Theta_{A}(s):[0,1]\rightarrow\mathbb{R} is the counterclockwise angle of the tangent vector \varphi^{\prime}(s) with respect to x -axis. When the curve is only piecewise C^{1} , which occurs for polygons, we can make the turning function well-defined