Defective Linear Layouts of Graphs
Abstract
A linear layout of a graph defines a total order of the vertices and partitions the edges into either stacks or queues, i.e., crossing-free and non-nested sets of edges along the order, respectively. In this work, we study defective linear layouts that allow forbidden patterns among edges of the same set. Our focus is on -defective stack layouts and -defective queue layouts, in which the conflict graph representing the forbidden patterns among the edges of each stack or queue has maximum degree at most .
Keywords and phrases:
Linear layouts, stack layouts, queue layouts, defective layoutsCategory:
Poster AbstractCopyright and License:
2012 ACM Subject Classification:
Mathematics of computing Graph algorithms ; Mathematics of computing Graph theoryAcknowledgements:
Research started at the Bertinoro Workshop on Graph Drawing 2024.Funding:
Bekos and Pavlidi were supported by HFRI Grant No: 26320.Editors:
Vida Dujmović and Fabrizio MontecchianiSeries and Publisher:
Leibniz International Proceedings in Informatics, Schloss Dagstuhl – Leibniz-Zentrum für Informatik
1 Introduction
Stack [3] and queue [5] layouts are key concepts in topological graph theory, where vertices are ordered and edges are partitioned into non-crossing (stacks) or non-nested (queues) sets. The goal is to minimize the number of stacks/queues – known as the stack/queue number. We introduce -defective linear layouts, a generalization of stack/queue layouts that allows some forbidden patterns (edge crossings or nestings), controlled by a conflict graph with maximum degree . Classical layouts correspond to the case . Our contributions include: (i) characterizations of -defective layouts; (ii) bounds on edge density; and (iii) bounds or exact values of -defective numbers for specific graph families.
2 Queue Layouts
We first bound the density of graphs admitting -defective -queue layouts. Let be a vertex ordering of a graph . Partition the edges into classes, where edges and belong to the same class if . We denote by the class whose maximum number of edges is (). Note that if has edges, it induces a path.
Lemma 1.
Let be a -defective -queue layout of a graph . Every defective queue of contains at most edges of each class.
Theorem 2.
Any -vertex graph that admits a -defective -queue layout has at most edges, for .
Sketch.
Let be a -defective -queue layout of a graph with vertices and edges. Classes with at most edges are small, otherwise are large. Let be the number of large classes, and the number of edges in small classes.
We claim . Suppose, for contradiction, that . Then some large class must have more than edges, implying a queue with more than edges from that class, contradicting Lemma 1.
Since each class has at most edges, we have , and . Therefore,
. Finally, for .
We next consider defectiveness and queue number. In this content, outerplanar graphs have queue number 2 [7], but not all admit -queue layouts with bounded defectiveness. Outer -planar graphs have queue number at most 42 [2]; we show that their -defective queue number is 2, so their queue number is at most 3.
For general planar graphs, we can prove an upper bound of on their -defective queue number by adapting a well-known technique by Dujmovic et al. [4].
We next establish bounds on the -defective queue number of that are tight for .
Theorem 3.
The -defective queue number of is at least and at most , where .
Sketch.
For the lower bound, consider a -defective queue layout of . The edges can be partitioned into classes . Since has edges and at most edges can share a defective queue (Lemma 1), at least queues are needed. We prove the upper bound via an explicit construction. For to be specified, assign to each defective queue (for ) all edges with hop-size for . The edges with the most nestings are the -hop edges, each nesting edges, ensuring the layout is -defective. To cover all edges of , note that each contains edges. For , the total edges assigned satisfy . We give bounds on the -defective queue number of , considering both the general case and the separated setting, where one part precedes the other [1].
Corollary 4.
The -defective queue number of in the separated setting is . For , it ranges between and in the separated setting, while in the non-separated setting it ranges between and , where .
3 Stack Layouts
A graph admits a -defective -stack layout if and only if its edges can be partitioned into defective stacks, each forming an outer -planar subgraph. This yields the following characterization.
Theorem 5.
A graph has -defective stack number if and only if it is outer -planar.
We leverage the characterization given above to obtain bounds on the edge density of the graphs admitting -defective -stack layouts.
Theorem 6 (Ábrego et al. [8], Pach et al. [6]).
Any -vertex graph that admits a -defective -stack layout has at most: (i) edges, if , (ii) edges, if , (iii) edges, if , (iv) , otherwise. Also, the first three bounds are tight.
Theorem 7.
An -vertex graph with a -defective -stack layout has at most edges.
In the following, we present bounds on the -defective stack numbers of and . For , our upper bounds are tight (up to small constants).
Theorem 8.
The -defective stack number of is at least and at most , where .
Corollary 9.
The -defective stack number of in the separated setting is at least and at most , while in the non-seperated setting it is at least and at most .
Theorem 10.
The -defective stack number of in the separated setting is at least and at most , where .
4 Open Problems
Our work raises several new open problems, which we list below.
-
Extend to other linear layouts (e.g., deques, riques) and mixed settings.
-
Explore defects beyond bounded conflict graph degree, such as bounding diameter.
-
Research questions from our results: (i) complexity of recognizing graphs with -defective -queue layouts for , (ii) improve bounds on -defective stack and queue numbers of for , and (iii) study other graph classes.
References
- [1] Jawaherul Md. Alam, Michael A. Bekos, Martin Gronemann, Michael Kaufmann, and Sergey Pupyrev. The mixed page number of graphs. Theor. Comput. Sci., 931:131–141, 2022. doi:10.1016/J.TCS.2022.07.036.
- [2] Michael A. Bekos, Martin Gronemann, and Chrysanthi N. Raftopoulou. An improved upper bound on the queue number of planar graphs. Algorithmica, 85(2):544–562, 2023. doi:10.1007/s00453-022-01037-4.
- [3] Frank Bernhart and Paul C. Kainen. The book thickness of a graph. J. Comb. Theory, Ser. B, 27(3):320–331, 1979. doi:10.1016/0095-8956(79)90021-2.
- [4] Vida Dujmović, Gwenaël Joret, Piotr Micek, Pat Morin, Torsten Ueckerdt, and David R. Wood. Planar graphs have bounded queue-number. J. ACM, 67(4):22:1–22:38, 2020. doi:10.1145/3385731.
- [5] Lenwood S. Heath and Arnold L. Rosenberg. Laying out graphs using queues. SIAM J. Comput., 21(5):927–958, 1992. doi:10.1137/0221055.
- [6] János Pach and Géza Tóth. Graphs drawn with few crossings per edge. Comb., 17(3):427–439, 1997. doi:10.1007/BF01215922.
- [7] S. Rengarajan and C. E. Veni Madhavan. Stack and queue number of 2-trees. In Ding-Zhu Du and Ming Li, editors, Computing and Combinatorics, First Annual International Conference, COCOON ’95, Xi’an, China, August 24-26, 1995, Proceedings, volume 959 of Lecture Notes in Computer Science, pages 203–212, 1995. doi:10.1007/BFB0030834.
- [8] Bernardo M. Ábrego, Joseph Dandurand, Silvia Fernández-Merchant, Emilie Lagoda, and Yury Sapozhnikov. Book crossing numbers of the complete graph and small local convex crossing numbers. CoRR, abs/1607.00131, 2024. URL: https://arxiv.org/abs/1607.00131.
