Abstract 1 Introduction 2 Background 3 Our contribution References

Reconfigurations of Plane Caterpillars and Paths

Todor Antić ORCID Department of Applied Mathematics, Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic Guillermo Gamboa Quintero ORCID Computer Science Institute of Charles University, Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic Jelena Glišić ORCID Department of Applied Mathematics, Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic
Abstract

Let S be a point set in the plane, and let 𝒫(S) and 𝒞(S) be the sets of all plane spanning paths and caterpillars on S. We study reconfiguration operations on 𝒫(S) and 𝒞(S). In particular, we prove that all of the commonly studied reconfigurations on plane spanning trees still yield connected reconfiguration graphs for caterpillars when S is in convex position. If S is in general position, we show that the rotation, compatible flip and flip graphs of 𝒞(S) are connected while the slide graph is sometimes disconnected, but always has a component of size 14(3n1). We then study sizes of connected components in reconfiguration graphs of plane spanning paths. In this direction, we show that no component of size at most 7 can exist in the flip graph on 𝒫(S).

Keywords and phrases:
reconfiguration graph, caterpillar, path, geometric graph
Category:
Poster Abstract
Funding:
Todor Antić: supported by grant no. 23-04949X of the Czech Science Foundation (GAČR) and by SVV–2023–260699.
Guillermo Gamboa Quintero: supported by the GAČR grant 22-17398S and the Charles University project PRIMUS/24/SCI/012.
Jelena Glišić: supported by grant no. 23-04949X of the Czech Science Foundation (GAČR) and by SVV–2023–260699.
Copyright and License:
[Uncaptioned image] © Todor Antić, Guillermo Gamboa Quintero, and Jelena Glišić; licensed under Creative Commons License CC-BY 4.0
2012 ACM Subject Classification:
Theory of computation Computational geometry
; Mathematics of computing Combinatorics
Related Version:
Full Version: https://arxiv.org/abs/2410.07419
Supplementary Material:
Acknowledgements:
We would like to thank Jan Kynčl and Pavel Valtr for proposing the problems and for all of the helpful discussions during the combinatorial problems seminar at Charles University. We also thank Daniel Perz for pointing out the existence of an isolated vertex in G𝒞slide(S), for his help in finding literature for this paper, and helpful comments.
Editors:
Vida Dujmović and Fabrizio Montecchiani

1 Introduction

Given a set of structures C, and a reconfiguration operation that transforms one object in C to another, the reconfiguration graph is a graph with vertex set C in which two vertices form an edge if one can be transformed into the other using a single reconfiguration operation. Often, in computer science, objects are solutions to a problem and reconfigurations are local changes that transform one solution into another. Then, to understand the solution space, it is important to study the reconfiguration graph. For an introduction to the topic, see [14].

Given a point set S in the plane, a plane spanning tree on S is a spanning tree of S whose edges are straight line segments that do not cross. Let 𝒯(S) be the set of all plane spanning trees on S. We consider five reconfigurations on 𝒯(S) (see Figure 1). For the following, we are given T1=(S,E1),T2=(S,E2)𝒯(S), and we say that:

  1. 1.

    T1 and T2 are connected by a flip if E2=E1{e}{f} for some edges e,f.

  2. 2.

    T1 and T2 are connected by a compatible flip if E2=E1{e}{f} for some edges e,f which do not cross.

  3. 3.

    T1 and T2 are connected by a rotation if E2=E1{e}{f} for some edges e,f which share an endpoint.

  4. 4.

    T1 and T2 are connected by an empty triangle rotation if E2=E1{e}{f} for some edges e,f which share an endpoint and the triangle spanned by their endpoints is empty.

  5. 5.

    T1 and T2 are connected by a slide if E2=E1{e}{f} for some edges e,f which are as in 4. and if e=ab and f=ac then bcE1E2.

Note that every slide is an empty triangle rotation, every empty triangle rotation is a rotation, and so on. This hierarchy will be useful when studying the structural properties of the corresponding reconfiguration graphs.

Refer to caption
Figure 1: a) A plane spanning tree. Replacing the dashed line with the dotted line corresponds to: b) a flip, c) a compatible flip, d) a rotation, e) an empty triangle rotation and f) a slide.

2 Background

The reconfiguration graphs associated with the operations described above have been a topic of interest for a long time with many results appearing through the years. These results have concerned connectivity [1, 7, 13], lower and upper bounds on the diameter [2, 9, 10], etc.

It is known that the reconfiguration graph of plane spanning trees is connected even in the most restrictive case when the reconfigurations are slides [1]. In the case of slides, a tight Θ(n2) bound on the diameter is shown in [4]. For the empty triangle rotations, an upper bound of O(nlogn) on the diameter was shown in [13], while for the remaining reconfigurations, a linear upper bound is known [7]. In [2], an upper bound of 2n3 is shown for flip graphs. The best known lower bound on the diameter of the reconfiguration graphs is 149nO(1), as per [8].

However, there has been little research on induced subgraphs of these reconfiguration graphs. The only such subgraph that has been explored is the one induced by plane spanning paths. For a set of n points in convex position, it is known that the flip graph is connected [5] and that it has diameter 2n5 for n{3,4} and 2n6 for n5 [11]. The flip graph of point sets with at most two convex layers is connected [12], as is the flip graph of generalized double circles [3]. For point sets in general position, connectivity was first conjectured by Akl, Islam and Meijer [5] 18 years ago but still remains unsolved.

3 Our contribution

We further the study of induced subgraphs of reconfiguration graphs of plane spanning trees by exploring reconfigurations of plane spanning caterpillars, and by expanding on the topic of reconfigurations of plane spanning paths. A caterpillar is a tree in which all non-leaf vertices form a path. We call this path the spine of the caterpillar. A plane spanning caterpillar of a point set S is a plane spanning tree of S which is a caterpillar. For a set S, we will denote by 𝒞(S) the set of all plane spanning caterpillars on S. We will denote the reconfiguration graphs on 𝒞(S) by G𝒞flip(S), G𝒞comp-flip(S), G𝒞rot(S), G𝒞emp-rot(S), and G𝒞slide(S). First, we focus on the case when S is in convex position. We show that the slide graph G𝒞slide is connected, which implies that all of the reconfiguration graphs are connected in this case. For compatible flips, we prove a stronger bound for the diameter.

Theorem 1.

Let S be a set of n3 points in convex position in the plane. Then, the graph G𝒞slide(S) is connected with diameter at most 3n8.

Theorem 2.

Let S be a set of n3 points in convex position in the plane. Then, the graph G𝒞comp-flip(S) is connected with diameter at most 2n5.

Next, we consider the case when S is a point set in general position. For a caterpillar C𝒞(S), and consecutive spine vertices vi,,vj of C, we write Si,j for the point set consisting of the spine vertices and all of the leaves attached to them. We call C𝒞(S) with spine v1,v2,,vk a well-separated caterpillar if for each i1, the convex hull of S1,i is disjoint from the rest of S. We can prove that all caterpillars in this class are mutually connected in G𝒞slide(S). This large connected component is significant, as we can also show that G𝒞slide(S) is disconnected. On the other hand, we can prove that the rotation graph G𝒞rot(S) is connected. We leave open the question of the connectivity of G𝒞emp-rot.

Theorem 3.

Any two well-separated caterpillars are connected in G𝒞slide(S).

Theorem 4.

The graph G𝒞slide(S) is connected for every set S of n points in the plane if n7. If n8, there exists a set S of n points such that G𝒞slide(S) has isolated vertices.

Proposition 5.

The graph G𝒞rot(S) is connected.

Lastly, we focus on connected components of the reconfiguration graph of plane spanning paths. Given a set of points in general position S, we will call the corresponding flip graph of plane spanning paths G𝒫(S). Currently, the main open problem is deciding whether G𝒫(S) is connected. Here we find a large connected component consisting of what we call generalized peeling paths. We note that Theorem 6 was independently discovered by Kleist, Kramer, and Rieck [12]. We use the number of these paths to prove that there are at least 14(3n1) well-separated caterpillars on S. Finally, we investigate the minimal size of components in G𝒫(S).

Theorem 6.

Let S be a set of n points in general position. Then G𝒫(S) contains a connected component of size 2Ω(n).

Theorem 7.

Let S be a point set of n5 points in general position. Then, G𝒫(S) contains no connected component of size at most 7.

References

  • [1] Oswin Aichholzer, Franz Aurenhammer, and Ferran Hurtado. Sequences of spanning trees and a fixed tree theorem. Computational Geometry, 21(1-2):3–20, 2002. doi:10.1016/S0925-7721(01)00042-6.
  • [2] Oswin Aichholzer, Brad Ballinger, Therese Biedl, Mirela Damian, Erik D. Demaine, Adam Hesterberg, Matias Korman, Anna Lubiw, Jayson Lynch, Josef Tkadlec, and et al. Reconfiguration of non-crossing spanning trees. Journal of Computational Geometry, 15(1):224–253, December 2024. doi:10.20382/jocg.v15i1a9.
  • [3] Oswin Aichholzer, Kristin Knorr, Wolfgang Mulzer, Johannes Obenaus, Rosna Paul, and Birgit Vogtenhuber. Flipping plane spanning paths. In WALCOM, volume 13973 of Lecture Notes in Computer Science, pages 49–60. Springer, 2023. doi:10.1007/978-3-031-27051-2_5.
  • [4] Oswin Aichholzer and Klaus Reinhardt. A quadratic distance bound on sliding between crossing-free spanning trees. Computational Geometry, 37(3):155–161, 2007. doi:10.1016/j.comgeo.2004.12.010.
  • [5] Selim G. Akl, Kamrul Islam, and Henk Meijer. On planar path transformation. Information Processing Letters, 104(2):59–64, 2007. doi:10.1016/j.ipl.2007.05.009.
  • [6] Todor Antić, Guillermo Gamboa Quintero, and Jelena Glišić. jelena-glisic/Caterpillar-Slides. Software, swhId: swh:1:dir:552a6b334147879c561c75db5d9b586bf6c23aa2 (visited on 2025-09-12). URL: https://github.com/jelena-glisic/Caterpillar-Slides, doi:10.4230/artifacts.24692.
  • [7] David Avis and Komei Fukuda. Reverse search for enumeration. Discrete Applied Mathematics, 65(1-3):21–46, 1996. doi:10.1016/0166-218X(95)00026-N.
  • [8] Håvard Bakke Bjerkevik, Linda Kleist, Torsten Ueckerdt, and Birgit Vogtenhuber. Flipping Non-Crossing Spanning Trees, pages 2313–2325. Society for Industrial and Applied Mathematics, 2025. doi:10.1137/1.9781611978322.77.
  • [9] Nicolas Bousquet, Lucas de Meyer, Théo Pierron, and Alexandra Wesolek. Reconfiguration of Plane Trees in Convex Geometric Graphs. In Wolfgang Mulzer and Jeff M. Phillips, editors, 40th International Symposium on Computational Geometry (SoCG 2024), volume 293 of Leibniz International Proceedings in Informatics (LIPIcs), pages 22:1–22:17, Dagstuhl, Germany, 2024. Schloss Dagstuhl – Leibniz-Zentrum für Informatik. doi:10.4230/LIPIcs.SoCG.2024.22.
  • [10] Nicolas Bousquet, Valentin Gledel, Jonathan Narboni, and Théo Pierron. A note on the flip distance between non-crossing spanning trees. Computating in Geometry and Topology, 2(1):8:1–8:7, 2023. doi:10.57717/cgt.v2i1.36.
  • [11] Jou-Ming Chang and Ro-Yu Wu. On the diameter of geometric path graphs of points in convex position. Information Processing Letters, 109(8):409–413, 2009. doi:10.1016/J.IPL.2008.12.017.
  • [12] Linda Kleist, Peter Kramer, and Christian Rieck. On the connectivity of the flip graph of plane spanning paths. In Daniel Kráľ and Martin Milanič, editors, Graph-Theoretic Concepts in Computer Science, pages 327–342, Cham, 2025. Springer Nature Switzerland. doi:10.1007/978-3-031-75409-8_23.
  • [13] Torrie L. Nichols, Alexander Pilz, Csaba D. Tóth, and Ahad N. Zehmakan. Transition operations over plane trees. Discrete Mathematics, 343(8):111929, 2020. doi:10.1016/j.disc.2020.111929.
  • [14] Naomi Nishimura. Introduction to reconfiguration. Algorithms, 11(4), 2018. doi:10.3390/a11040052.