DOCSLIB.ORG

Interval Orders

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text

Load more

Interval Orders Combinatorial Structure and Algorithms im Februar von Dipl Math Stefan Felsner am Fachb ereich Mathematik der Technischen Universitat Berlin vorgelegte Dissertation D Preface This thesis is based on my research on partially ordered sets and sp ecially inter val orders that b egan when I came to Berlin in Professor RH Mohring intro duced me to the eld I like to thank him for stimulations and guidance over the years Beside the intro duction the thesis combines ve chapters that have in common the central role played by interval orders On the other hand they are only lo osely connected and so I decided to make the chapters selfcontained To emphasize the indep endency of the chapters references are given at the end of each one Articles and b o oks that are of signicance to the general theme and have b een consulted without b eing cited directly are collected in the references of the intro duction An outline of the contents of the thesis can b e found in the preview at the b eginning of the intro duction All further chapters are op ened by a section called Intro duction and Overview That sp ecial section may serve as an extended abstract for the contents of the chapter it also gives the relationship to the existing literature I am indebted to many p eople for encouragement and discussions Sp ecial thanks go to Tom Trotter Lorenz Wernisch and the memb ers of our group dis crete algorithmical mathematics which provided an ageeable and creative atmo sphere Stefan Felsner Berlin February Contents Intro duction Preview Partially Ordered Sets Interval Orders References for Chapter The Jump Numb er of Interval Orders Intro duction and Overview Greedy Linear Extensions and Starting Elements Bounds for the Jump Numb er of Interval Orders An Approximation Algorithm Defect Optimal Interval Orders The NP Completeness Pro of References for Chapter The Dimension of Interval Orders Intro duction and Overview Some Logarithmic Bounds Marking Functions and Bounds The StepGraph and More Bounds Op en Problems The Doubly Logarithmic Bound Canonical Interval Orders General Interval Orders Colorings of Digraphs and ShiftGraphs Colorings of Directed Graphs Asymptotics CONTENTS A Connection with Lattices of Antichains References for Chapter Coloring Interval Order Diagrams Intro duction and Overview A Construction of Long Sequences The Structure of Very Long Sequences Long Cycles b etween Consecutive Levels in B n References for Chapter Interval Dimension and Dimension Intro duction and Overview Reducing Interval Dimension to Dimension Consequences Crowns and the Standard Examples B and the Split Op eration The Interval Dimension of Sub divisions Comparability Invariance of Interval Dimension Remarks on Computational Complexity References for Chapter Tolerance Graphs Intro duction and Overview Tolerance Graphs and Interval Dimension Some Examples Cotrees and More Examples References for Chapter Chapter Intro duction Preview Here we only give a very brief survey of the contents of this thesis For more detailed information on the sub ject of a single chapter we refer to intro ductory sections called Intro duction and Overview there we also relate our own results to the existing literature A family I of intervals on the real line may b e used to dene a partial x xX ordering on X We put x y if a I and b I implies a b A partial order x y I X thus obtained is called an interval order In this chapter we intro duce basic terminology and facts ab out partial orders linear extensions and interval orders that are used throughout all parts The next three chapters then deal with problems for interval orders that are known to b e NP complete for general partial orders In the next chapter we consider the jump number problem That is we are lo oking for a linear extension of I which has a minimal numb er of adjacent pairs that are incomparable in I We intro duce parameters and of an interval order and prove a close relationship with the jump numb er The jump numb er of an interval order is at least max A linear extension with at most jumps can always b e found This is an approximation ratio of There is a p olynomial algorithm which decides whether the jump numb er is exactly It is NP complete to decide whether the jump numb er is exactly Chapter deals with the dimension of interval orders The dimension of a partial CHAPTER INTRODUCTION order is the minimal dimension of an euklidean space that admits an emb edding of the order We rst use the concept of marking intervals to obtain easy pro ofs for some logarithmic b ounds Afterwards we intro duce the stepgraph of an interval order A partition of the arcs of the stepgraph into semitransitive classes leads to a realizer This observation is used to give more b ounds particularly a doubly logarithmic b ound Motivated by the pro of of the lower b ound for the dimension of interval orders we nally study colorings and arccolorings of digraphs and stepgraphs In Chapter we directly deal with colorings namely with the chromatic num b er of the diagram of an interval order It is shown that log n colors always suce for diagrams of hight n On the other hand there are interval orders of this hight such that the diagram requires log n colors The construction of go o d colorings relies on the existence of certain sequences of sets of colors sequences An upp er b ound for the length of sequences is given and we show that sequences of this length corresp ond to level accurate Hamiltonian paths in the Bo olean lattice In Bo olean lattices of order we could construct such paths Chapter deals with the interplay of interval dimension and dimension We dene a transformation P Q b etween partial orders such that the dimension of Q and the interval dimension of P agree We provide two interpretations of this transformation a combinatorial one and a geometrical one These two in terpretations are used for several consequences We relate the interval dimension of sub divisions of an order P to the dimension of P We give a new pro of for the comparability invariance of interval dimension Finally in the last chapter we deal with tolerance graphs The complement of a b ounded tolerance graph has an orientation as partial order of interval di mension This order admits a square representation while general orders of interval dimension require nonsquare rectangels in their representation This observation is the starting p oint for our invertigations ab out the relations b e tween several classes of graphs Some of our results are If the complement of a tolerance graph admits an orientation as partial order then this order has inter val dimension We give an example of an alternatingly orientable graph that is the complement of an order of interval dimension but is not a tolerance graph We also characterize the tolerance graphs among the complements of trees PARTIALLY ORDERED SETS Partially Ordered Sets In this and the next section we intro duce necessary terminology and basic results ab out partially ordered sets linear extensions and interval orders A strict partial ordering R on a nite set X is a binary relation on X such that x x is not in R for any x X if x y and y z are in R then x z is in R Condition is the irreexivity of R and Condition says that R is transitive Note that as a consequence of the two conditions we obtain that R is antisym metric ie if x y R and x y then y x is not in R In some situations it is appropriate to use the idea of a reexive partial ordering This is a relation R which is reexive ie x x is in R for each x X antisymmetric and transitive The dierence b etween a strict and a reexive ordering is normally clear from the context and the word ordering may refer to either kind of ordering A set X together with a partial ordering R on X is called a partial ly ordered set and denoted by X R We often abbreviate partially ordered set as either ordered set order or poset The commonly used notation for a partially ordered set P is P X in the case of a strict ordering and P X if the ordering is reexive We then write x y instead of x y and y x to mean x y There is a further abuse of notation which should b e mentioned If P X is a p oset we sometimes write x P or x y P instead of x X and x y The comparability graph of a p oset P X denoted CompP is a graph on X its edges are the comparabilities of P ie fx yg is an edge if either x y or y x A binary relation R on a set X can b e represented graphically by a directed graph digraph for short We represent the elements of X as p oints and use arrows arcs to represent the ordered pairs in R When the binary
Recommended publications
  • Confluent Hasse Diagrams

    Confluent Hasse Diagrams

    Confluent Hasse diagrams David Eppstein Joseph A. Simons Department of Computer Science, University of California, Irvine, USA. October 29, 2018 Abstract We show that a transitively reduced digraph has a confluent upward drawing if and only if its reachability relation has order dimension at most two. In this case, we construct a confluent upward drawing with O(n2) features, in an O(n) × O(n) grid in O(n2) time. For the digraphs representing series-parallel partial orders we show how to construct a drawing with O(n) fea- tures in an O(n) × O(n) grid in O(n) time from a series-parallel decomposition of the partial order. Our drawings are optimal in the number of confluent junctions they use. 1 Introduction One of the most important aspects of a graph drawing is that it should be readable: it should convey the structure of the graph in a clear and concise way. Ease of understanding is difficult to quantify, so various proxies for readability have been proposed; one of the most prominent is the number of edge crossings. That is, we should minimize the number of edge crossings in our drawing (a planar drawing, if possible, is ideal), since crossings make drawings harder to read. Another measure of readability is the total amount of ink required by the drawing [1]. This measure can be formulated in terms of Tufte’s “data-ink ratio” [22,35], according to which a large proportion of the ink on any infographic should be devoted to information. Thus given two different ways to present information, we should choose the more succinct and crossing-free presentation.
  • Interval Reductions and Extensions of Orders: Bijections To

    Interval Reductions and Extensions of Orders: Bijections To

    Interval Reductions and Extensions of Orders Bijections to Chains in Lattices Stefan Felsner Jens Gustedt and Michel Morvan Freie Universitat Berlin Fachbereich Mathematik und Informatik Takustr Berlin Germany Email felsnerinffub erlinde Technische Universitat Berlin Fachbereich Mathematik Strae des Juni MA Berlin Germany Email gustedtmathtub erlinde LIAFA Universite Paris Denis Diderot Case place Jussieu Paris Cedex France Email morvanliafajussieufr Abstract We discuss bijections that relate families of chains in lattices asso ciated to an order P and families of interval orders dened on the ground set of P Two bijections of this typ e havebeenknown The bijection b etween maximal chains in the antichain lattice AP and the linear extensions of P A bijection between maximal chains in the lattice of maximal antichains A P and minimal interval extensions of P M We discuss two approaches to asso ciate interval orders to chains in AP This leads to new bijections generalizing Bijections and As a consequence wechar acterize the chains corresp onding to weakorder extensions and minimal weakorder extensions of P Seeking for a way of representing interval reductions of P bychains wecameup with the separation lattice S P Chains in this lattice enco de an interesting sub class of interval reductions of P Let S P b e the lattice of maximal separations M in the separation lattice Restricted to maximal separations the ab ove bijection sp ecializes to a bijection whic h nicely complementsand A bijection between maximal
  • Dyck Paths and Positroids from Unit Interval Orders (Extended Abstract)

    Dyck Paths and Positroids from Unit Interval Orders (Extended Abstract)

    Dyck Paths and Positroids from Unit Interval Orders (Extended Abstract) Anastasia Chavez 1 and Felix Gotti 2 ∗ 1Department of Mathematics, UC Berkeley, Berkeley CA 94720 2 Department of Mathematics, UC Berkeley, Berkeley CA 94720 . Abstract. It is well known that the number of non-isomorphic unit interval orders on [n] equals the n-th Catalan number. Using work of Skandera and Reed and work of Postnikov, we show that each unit interval order on [n] naturally induces a rank n positroid on [2n]. We call the positroids produced in this fashion unit interval positroids. We characterize the unit interval positroids by describing their associated decorated permutations, showing that each one must be a 2n-cycle encoding a Dyck path of length 2n. 1 Introduction A unit interval order is a partially ordered set that captures the order relations among a collection of unit intervals on the real line. Unit interval orders were introduced by Luce [8] to axiomatize a class of utilities in the theory of preferences in economics. Since then they have been systematically studied (see [3,4,5,6, 14] and references therein). These posets exhibit many interesting properties; for example, they can be characterized as the posets that are simultaneously (3 + 1)-free and (2 + 2)-free. Moreover, it is well 1 2n known that the number of non-isomorphic unit interval orders on [n] equals n+1 ( n ), the n-th Catalan number (see [3, Section 4] or [15, Exercise 2.180]). In [14], motivated by the desire to understand the f -vectors of various classes of posets, Skandera and Reed showed that one can canonically label the elements of a unit interval order from 1 to n so that its n × n antiadjacency matrix is totally nonnegative (i.e., has all its minors nonnegative) and its zero entries form a right-justified Young diagram located strictly above the main diagonal and anchored in the upper-right corner.
  • LNCS 7034, Pp

    LNCS 7034, Pp

    Confluent Hasse Diagrams DavidEppsteinandJosephA.Simons Department of Computer Science, University of California, Irvine, USA Abstract. We show that a transitively reduced digraph has a confluent upward drawing if and only if its reachability relation has order dimen- sion at most two. In this case, we construct a confluent upward drawing with O(n2)features,inanO(n) × O(n)gridinO(n2)time.Forthe digraphs representing series-parallel partial orders we show how to con- struct a drawing with O(n)featuresinanO(n)×O(n)gridinO(n)time from a series-parallel decomposition of the partial order. Our drawings are optimal in the number of confluent junctions they use. 1 Introduction One of the most important aspects of a graph drawing is that it should be readable: it should convey the structure of the graph in a clear and concise way. Ease of understanding is difficult to quantify, so various proxies for it have been proposed, including the number of crossings and the total amount of ink required by the drawing [1,18]. Thus given two different ways to present information, we should choose the more succinct and crossing-free presentation. Confluent drawing [7,8,9,15,16] is a style of graph drawing in which multiple edges are combined into shared tracks, and two vertices are considered to be adjacent if a smooth path connects them in these tracks (Figure 1). This style was introduced to re- duce crossings, and in many cases it will also Fig. 1. Conventional and confluent improve the ink requirement by represent- drawings of K5,5 ing dense subgraphs concisely.
  • Degree Bounds for Linear Discrepancy of Interval Orders And

    Degree Bounds for Linear Discrepancy of Interval Orders And

    A Brooks-Type Theorem for the Bandwidth of Interval Graphs Mitchel T. Kellera, Stephen J. Youngb aSchool of Mathematics, Georgia Institute of Technology, Atlanta, Georgia 30332-0160 U.S.A. bSchool of Mathematics, Georgia Institute of Technology, Atlanta, Georgia 30332-0160 U.S.A. Abstract Let G be an interval graph. The layout that arranges the intervals in order by right endpoint easily shows that the bandwidth of G is at most its maximum degree ∆. Hence, if G contains a clique of size ∆ + 1, then its bandwidth must be ∆. In this paper we prove a Brooks-type bound on the bandwidth of interval graphs. Namely, the bandwidth of an interval graph is at most ∆, with equality if and only if it contains a clique of size ∆+1. Furthermore, the stronger bound is tight even for interval graphs of clique number 2. Our proof utilizes the correspondence between the linear discrepancy of a partially ordered set and the bandwidth of its co-comparability graph. We also make progress toward a related question of Tanenbaum, Trenk, and Fishburn. They asked if a poset in which each point is incomparable to at most ∆ others has linear discrepancy at most b(3∆ − 1)=2c. We show that this is true if the poset is disconnected. Key words: bandwidth, interval graph, linear discrepancy, interval order, poset 2000 MSC: 05C78, 05C62, 06A07 1. Introduction The bandwidth of a graph G on n vertices is the least integer k such that there exists a labelling of the vertices by [n] = f1; 2; : : : ; ng so that the labels of adjacent vertices differ by at most k.
  • Degree Bounds for Linear Discrepancy of Interval Orders and Disconnected Posets

    Degree Bounds for Linear Discrepancy of Interval Orders and Disconnected Posets

    Degree Bounds for Linear Discrepancy of Interval Orders and Disconnected Posets Mitchel T. Kellera, Stephen J. Younga,1 aSchool of Mathematics, Georgia Institute of Technology, Atlanta, Georgia 30332-0160 U.S.A. Abstract Let P be a poset in which each point is incomparable to at most D others. Tanenbaum, Trenk, and Fishburn asked in a 2001 paper if the linear discrepancy of such a poset is bounded above by b(3D − 1)=2c. This paper answers their question in the affirmative for two classes of posets. The first class is the interval orders, which are shown to have linear discrepancy at most D, with equality precisely for interval orders containing an antichain of size D + 1. The stronger bound is tight even for interval orders of width 2. The second class of posets considered is the dis- connected posets, which have linear discrepancy at most b(3D − 1)=2c. This paper also contains lemmas on the role of critical pairs in linear discrepancy as well as a theorem establishing that every poset contains a point whose removal decreases the linear discrepancy by at most 1. Key words: linear discrepancy, interval order, poset, bandwidth, interval graph 2010 MSC: 06A07, 05C78, 05C62 1. Introduction Tanenbaum, Trenk, and Fishburn introduced the concept of linear discrepancy in [15] as a way of measuring how far apart linear extensions must place incomparable elements. They closed that paper with eight questions and challenges. One called for special results for linear dis- crepancy of interval orders, and another asked a question about bounding the linear discrepancy of a poset in terms of the maximum number of elements with which any element is incomparable.
  • The Complexity of the Partial Order Dimension Problem-Closing The

    The Complexity of the Partial Order Dimension Problem-Closing The

    The Complexity of the Partial Order Dimension Problem – Closing the Gap † Stefan Felsner∗ Irina Mustat¸a˘ Martin Pergel Department of Software and Institut f¨ur Mathematik Berlin Mathematical School Computer Science Education Technische Universit¨at Berlin Technische Universit¨at Berlin Charles University Strasse des 17. Juni 136 Strasse des 17. Juni 136 Malostransk´en´am. 25 D-10623 Berlin, Germany D-10623 Berlin, Germany 118 00 Praha 1, Czech Republic Abstract The dimension of a partial order P is the minimum number of linear orders whose intersection is P . There are efficient algorithms to test if a partial order has dimension at most 2. In 1982 Yannakakis [25] showed that for k ≥ 3 to test if a partial order has dimension ≤ k is NP-complete. The height of a partial order P is the maximum size of a chain in P . Yannakakis also showed that for k ≥ 4 to test if a partial order of height 2 has dimension ≤ k is NP-complete. The complexity of deciding whether an order of height 2 has dimension 3 was left open. This question became one of the best known open problems in dimension theory for partial orders. We show that the problem is NP-complete. Technically, we show that the decision problem (3DH2) for dimension is equivalent to deciding for the existence of bipartite triangle containment representations (BTCon). This problem then allows a reduction from a class of planar satisfiability problems (P-3-CON-3-SAT(4)) which is known to be NP-hard. Mathematics Subject Classifications (2010) 06A07, 68Q25, 05C62, Keywords : NP-complete, triangle containment, intersection graph.
  • Satomi Okazaki

    Satomi Okazaki

    Cycle Types of Permutations with Restricted Positions and a Characterization of a New Class of Interval Orders by Satomi Okazaki B.S. Mathematics California Institute of Technology, 1990 Submitted to the Department of Mathematics in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Mathematics at the Massachusetts Institute of Technology June 1996 @ Massachusetts Institute of Technology, 1996 All rights reserved. Signature of Author DIpartment of Mathematics May 3 1996 Certified by Richard P tanley Professor of Applied M thematics •4esis Supervi-,- Accepted by Richard P.$anley Chairman, Applied Mathematics (*rmmittee Accepted by 15avid A. Vogan, Chairman Departmental Graduate Committee Department of Mathematics OF T41.7,(3N L JUL 0 8 1996 Sc~e ~ Cycle Types of Permutations with Restricted Positions and a Characterization of a New Class of Interval Orders by Satomi Okazaki Submitted to the Department of Mathematics on May 3, 1996, in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Mathematics ABSTRACT This thesis deals with two unrelated problems. In Chapter 1, we examine permutations with restricted positions according to cycle type. R. Stanley and J. Stembridge developed this theory and defined "cycle indicators" to attack some conjectures about immanants of the Jacobi-Trudi matrix. They also proved a general formula for computing these cycle indicators. We calcu- late some cycle indicators using a different method for particular restrictions and prove some nice identities along the way. We also generalize the theory of classifying permutations in S,, with restricted positions by cycle type to wreath products of finite abelian groups with the symmetric group.
  • Convexity in Partial Cubes: the Hull Number

    Convexity in Partial Cubes: the Hull Number

    Convexity in partial cubes: the hull number Marie Albenque1 and Kolja Knauer2 1 LIX UMR 7161, Ecole´ Polytechnique, CNRS, France 2 LIF UMR 7279, Universit´eAix-Marseille, CNRS, France Abstract. We prove that the combinatorial optimization problem of de- termining the hull number of a partial cube is NP-complete. This makes partial cubes the minimal graph class for which NP-completeness of this problem is known and improves earlier results in the literature. On the other hand we provide a polynomial-time algorithm to determine the hull number of planar partial cube quadrangulations. Instances of the hull number problem for partial cubes described include poset dimension and hitting sets for interiors of curves in the plane. To obtain the above results, we investigate convexity in partial cubes and obtain a new characterization of these graphs in terms of their lattice of convex subgraphs. This refines a theorem of Handa. Furthermore we provide a topological representation theorem for planar partial cubes, generalizing a result of Fukuda and Handa about tope graphs of rank 3 oriented matroids. 1 Introduction The objective of this paper is the study of convexity and particularly of the hull number problem on different classes of partial cubes. Our contribution is twofold. First, we establish that the hull number problem is NP-complete for partial cubes, second, we emphasize reformulations of the hull number problem for certain classes of partial cubes leading to interesting problems in geometry, poset theory and plane topology. In particular, we provide a polynomial time algorithm for the class of planar partial cube quadrangulations.
  • Maximal Cliques Structure for Cocomparability Graphs and Applications

    Maximal Cliques Structure for Cocomparability Graphs and Applications

    Maximal cliques structure for cocomparability graphs and applications J´er´emieDusart∗, Michel Habib∗yand Derek G. Corneilz July 6, 2018 Abstract A cocomparability graph is a graph whose complement admits a transitive orien- tation. An interval graph is the intersection graph of a family of intervals on the real line. In this paper we investigate the relationships between interval and cocomparabil- ity graphs. This study is motivated by recent results [5, 13] that show that for some problems, the algorithm used on interval graphs can also be used with small modifica- tions on cocomparability graphs. Many of these algorithms are based on graph searches that preserve cocomparability orderings. First we propose a characterization of cocomparability graphs via a lattice structure on the set of their maximal cliques. Using this characterization we can prove that every maximal interval subgraph of a cocomparability graph G is also a maximal chordal subgraph of G. Although the size of this lattice of maximal cliques can be exponential in the size of the graph, it can be used as a framework to design and prove algorithms on cocomparability graphs. In particular we show that a new graph search, namely Local Maximal Neighborhood Search (LocalMNS) leads to an O(n + mlogn) time algorithm to find a maximal interval subgraph of a cocomparability graph. Similarly we propose a linear time algorithm to compute all simplicial vertices in a cocomparability graph. In both cases we improve on the current state of knowledge. Keywords: (co)-comparability graphs, interval graphs, posets, maximal antichain lat- arXiv:1611.02002v1 [cs.DM] 7 Nov 2016 tices, maximal clique lattices, graph searches.
  • The Interval Inclusion Number of a Partially Ordered Set

    The Interval Inclusion Number of a Partially Ordered Set

    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Discrete Mathematics 88 (1991) 259-277 259 North-Holland The interval inclusion number of a partially ordered set Tom Madej* Department of Computer Science, University of Illinois, Urbana, IL 61801, USA Douglas B. West* * Department of Mathematics, University of Illinois, Urbana, IL 61801, USA Received 13 September 1988 Abstract Madej, T. and D. B. West, The interval inclusion number of a partially ordered set, Discrete Mathematics 88 (1991) 259-277. A containment representation for a poset P is a map f such that x < y in P if and only if f(x) cf(y). We introduce the interval inclusion number (or interval number) i(P) as the smallest t such that P has a containment representation fin which each f(x) is the union of at most t intervals. Trivially, i(P) = 1 if and only if dim(P) =Z2. Posets with i(P) = 2 include the standard n-dimensional poset and all interval orders; i.e. posets of arbitrarily high dimension. In general we have the upper bound i(P) < [dim(P)/2], with equality holding for the Boolean algebras. For lexicographic composition, i(P) = k and dim(Q) = 2k + 1 imply i(P[Q]) = k + 1. This result and i(&J = k imply that testing i(P) 6 k for any fixed k 2 2 is NP-complete. Concerning removal theorems, we prove that i(P -x) 5 i(P) - 1 when x is a maximal or minimal element of P, and in general i(P - x) 2 i(P)/2.
  • Upright-Quad Drawing of St-Planar Learning Spaces David Eppstein Computer Science Department, University of California, Irvine Eppstein@Uci.Edu

    Upright-Quad Drawing of St-Planar Learning Spaces David Eppstein Computer Science Department, University of California, Irvine [email protected]

    Journal of Graph Algorithms and Applications http://jgaa.info/ vol. 12, no. 1, pp. 51–72 (2008) Upright-Quad Drawing of st-Planar Learning Spaces David Eppstein Computer Science Department, University of California, Irvine [email protected] Abstract We consider graph drawing algorithms for learning spaces, a type of st-oriented partial cube derived from an antimatroid and used to model states of knowledge of students. We show how to draw any st-planar learning space so all internal faces are convex quadrilaterals with the bot- tom side horizontal and the left side vertical, with one minimal and one maximal vertex. Conversely, every such drawing represents an st-planar learning space. We also describe connections between these graphs and arrangements of translates of a quadrant. Our results imply that an an- timatroid has order dimension two if and only if it has convex dimension two. Article Type Communicated by Submitted Revised Regular paper M. Kaufmann and D. Wagner December 2006 February 2008 A preliminary version of this paper was presented in 2006 at the 14th International Symposium on Graph Drawing in Karlsruhe, Germany. A condensed treatment of this material also appears in Section 11.6 of [10]. Eppstein, Upright-Quad Drawing, JGAA, 12(1) 51–72 (2008) 52 1 Introduction A partial cube is a graph that can be given the geometric structure of a hyper- cube, by assigning the vertices bitvector labels in such a way that the graph distance between any pair of vertices equals the Hamming distance of their la- bels. Partial cubes can be used to describe benzenoid systems in chemistry [13], weak or partial orderings modeling voter preferences in multi-candidate elections [12], integer partitions in number theory [8], flip graphs of point set triangulations [9], and the hyperplane arrangements familiar to computational geometers [15, 7]; see [10] for additional applications.