arXiv subjects

Maurice Pouzet

Publications and source records attributed to Maurice Pouzet.

At least 19 recordsLinked to original sources

arXiv2026

The structure of interval orders with no infinite antichain

We prove that if $G=(V,E)$ is a nonprime graph with either no infinite independent set or no infinite clique, then every vertex of $G$ belongs to a maximal strong module distinct from $V$. In particular, $G$ admits a Gallai decomposition. As a consequence, we obtain that every interval order $P$ with no infinite antichain admits a Gallai decomposition. That is, $P$ is a lexicographical sum of interval orders distinct from $P$ indexed by either a chain, an antichain, or a prime interval order. Next, we prove that every prime interval order with no infinite antichain is at most countable and does not embed a copy of the chain of rational numbers. Finally, for each countable ordinal $α$, we construct a well-quasi-ordered prime interval order $P_α$ whose chain of maximal antichains has Hausdorff rank $α$.

math.CO↗

arXiv2025

On the ABK Conjecture, alpha-well Quasi Orders and Dress-Schiffels product

The following is a 2008 conjecture of Abraham, Bonnet and Kubiś: [ABK Conjecture] Every well quasi order (wqo) is a countable union of better quasi orders (bqo). We obtain a partial progress on the conjecture, by showing that the class of orders that are a countable union of better quasi orders (sigma-bqo) is closed under various operations. These include diverse products, such as the Dress-Shieffels product. We develop various properties of the latter product. In relation with the main question, we explore the class of alpha-wqo for countable ordinals alpha and obtain several closure properties and a Hausdorff-style classification theorem. Our main contribution is the discovery of various properties of sigma-bqos and ruling out potential counterexamples to the ABK Conjecture.

math.LO↗

arXiv2023

Absolute retracts of reflexive oriented graphs: the role of the MacNeille completion

We characterize the absolute retracts in the category of reflexive oriented graphs, that is, antisymmetric reflexive graphs, where morphisms between objects preserve arcs (which may be sent to loops). Here we show, by correcting a much earlier attempt at a proof, that a reflexive oriented graph is an absolute retract if and only if it is indeed a retract of some (direct) product of reflexive oriented zigzags (which are concatenations of reflexive oriented paths). Absolute retracts are therefore necessarily acyclic. In contrast to other categories of graphs and ordered sets, not every acyclic oriented graph can be embedded isometrically into some absolute retract. Embedding involves isometry with respect to the zig-zag distances forming a particular "dual quantale", which is a complete lattice of certain sets of words over the alphabet $\{+, -\}$, endowed with an additional monoid operation (viz., compound concatenation of sets of words) and an involution (interchanging $+$ and $- $ and then mirroring words). As reflexive oriented zigzags have MacNeille-closed distances, so do their products and retracts. So, the category of reflexive oriented graphs and its full subcategory of reflexive acyclic graphs do not have enough injectives, as the injective objects coincide with the absolute retracts.

math.CO↗

arXiv2023

Ordered structures with no finite monomorphic decomposition. Application to the profile of hereditary classes

We present a structural approach of some results about jumps in the behavior of the profile (alias generating function) of hereditary classes of finite structures. We consider the following notion due to N.Thiéry and the second author. A \emph{monomorphic decomposition} of a relational structure $R$ is a partition of its domain $V(R)$ into a family of sets $(V_x)_{x\in X}$ such that the restrictions of $R$ to two finite subsets $A$ and $A'$ of $V(R)$ are isomorphic provided that the traces $A\cap V_x$ and $A'\cap V_x$ have the same size for each $x\in X$. Let $\mathscr S_μ$ be the class of relational structures of signature $μ$ which do not have a finite monomorphic decomposition. We show that if a hereditary subclass $\mathscr D$ of $\mathscr S_μ$ is made of ordered relational structures then it contains a finite subset $\mathfrak A$ such that every member of $\mathscr D$ embeds some member of $\mathfrak A$. Furthermore, for each $R\in \mathfrak A$ the profile of the age $\age(R)$ of $R$ (made of finite substructures of $R$) is at least exponential. We deduce that if the profile of a hereditary class of finite ordered structures is not bounded above by a polynomial then it is at least exponential. For ordered graphs, this result is a part of classification obtained by Balogh, Bollobás and Morris (2006).

math.CO↗

arXiv2023

A few characterizations of topological spaces with no infinite discrete subspace

We give several characteristic properties of FAC spaces, namely topological spaces with no infinite discrete subspace. The first one was obtained in 2019 by the first author, and states that every closed set is a finite union of irreducible closed subsets. The full result extends well-known characterizations of posets with no infinite antichain. One of them is that FAC spaces are, equivalently, topological spaces in which every closed set contains a dense Noetherian subspace, or spaces in which every Hausdorff subspace is finite, or in which no subspace has any infinite relatively Hausdorff subset. The latter comes with a nice min-max property, extending an observation of Erdös and Tarski in the case of posets: on spaces with no infinite relatively Hausdorff subset, the cardinalities of relatively Hausdorff subsets are bounded, and the least upper bound is also the least cardinality of a family of closed irreducible subsets that cover the space.

math.GN↗

arXiv2022

The chain covering number of a poset with no infinite antichains

The chain covering number $\Cov(P)$ of a poset $P$ is the least number of chains needed to cover $P$. For a cardinal $ν$, we give a list of posets of cardinality and covering number $ν$ such that for every poset $P$ with no infinite antichain, $\Cov(P)\geq ν$ if and only if $P$ embeds a member of the list. This list has two elements if $ν$ is a successor cardinal, namely $[ν]^2$ and its dual, and four elements if $ν$ is a limit cardinal with $\cf(ν)$ weakly compact. For $ν= \aleph_1$, a list was given by the first author; his construction was extended by F. Dorais to every infinite successor cardinal $ν$.

math.CO↗

arXiv2022

On the Boolean dimension of a graph and other related parameters

We present the Boolean dimension of a graph, we relate it with the notions of inner, geometric and symplectic dimensions, and with the rank and minrank of a graph. We obtain an exact formula for the Boolean dimension of a tree in terms of a certain star decomposition. We relate the Boolean dimension with the inversion index of a tournament.

math.CO↗

arXiv2022

Hereditary classes of ordered sets of width at most two

This paper is a contribution to the study of hereditary classes of relational structures, these classes being quasi-ordered by embeddability. It deals with the specific case of ordered sets of width two and the corresponding bichains and incomparability graphs. Several open problems about hereditary classes of relational structures which have been considered over the years have positive answer in this case. For example, well-quasi-ordered hereditary classes of finite bipartite permutation graphs, respectively finite 321-avoiding permutations, have been characterized by Korpelainen, Lozin and Mayhill, respectively by Albert, Brignall, Ruškuc and Vatter. We provide another proof of the results mentioned above. It is based on the existence of a countable universal poset of width two, obtained by the first author in 1978, his notion of multichainability (1978) (a kind of analog to letter-graphs), and metric properties of incomparability graphs. Using Laver's theorem (1971) on better-quasi-ordering (bqo) of countable chains we prove that a wqo hereditary class of finite or countable bipartite permutation graphs is necessarily bqo. This gives a positive answer to a conjecture of Nash-Williams (1965) in this case. We extend a previous result of Albert et al. by proving that if a hereditary class of finite, respectively countable, bipartite permutation graphs is wqo, respectively bqo, then the corresponding hereditary classes of posets of width at most two and bichains are wqo, respectively bqo. Several notions of labelled wqo are also considered. We prove that they are all equivalent in the case of bipartite permutation graphs, posets of width at most two and the corresponding bichains. We characterize hereditary classes of finite bipartite permutation graphs which remain wqo when labels from a wqo are added.

math.CO↗

arXiv2022

Minimal prime ages, words and permutation graphs

This paper is a contribution to the study of hereditary classes of finite graphs. We classify these classes according to the number of prime structures they contain. We consider such classes that are \emph{minimal prime}: classes that contain infinitely many primes but every proper hereditary subclass contains only finitely many primes. We give a complete description of such classes. In fact, each one of these classes is a well-quasi-ordered (w.q.o) age and there are uncountably many of them. Eleven of these ages are almost multichainable; they remain w.q.o when labels in a w.q.o are added, hence have finitely many bounds. Five ages among them are exhaustible. Among the remaining ones, only countably many remain w.q.o when one label is added, and these have finitely many bounds (except for the age of the infinite path and its complement). The others have infinitely many bounds. Except for six examples, members of these ages we characterize are permutation graphs. In fact, every age which is not among the eleven ones is the age of a graph associated to a uniformly recurrent word on the integers. A description of minimal prime classes of posets and bichains is also provided. Our results support the conjecture that if a hereditary class of finite graphs does not remain w.q.o when adding labels from a w.q.o set to these graphs, then it is not w.q.o if we add just two constants to each of these graphs Our description of minimal prime classes uses a description of minimal prime graphs \cite{pouzet-zaguia2009} and previous work by Sobrani \cite{sobranithesis, sobranietat} and the authors \cite{oudrar, pouzettr} on properties of uniformly recurrent words and the associated graphs. The completeness of our description is based on classification results of Chudnovsky, Kim, Oum and Seymour \cite{chudnovsky} and Malliaris and Terry \cite {malliaris}.

math.CO↗

arXiv2022

Minimal prime ages, words and permutation graphs Extended abstract

math.CO↗

arXiv2022

Graphs containing finite induced paths of unbounded length

The age $\mathcal{A}(G)$ of a graph $G$ (undirected and without loops) is the collection of finite induced subgraphs of $G$, considered up to isomorphy and ordered by embeddability. It is well-quasi-ordered (wqo) for this order if it contains no infinite antichain. A graph is \emph{path-minimal} if it contains finite induced paths of unbounded length and every induced subgraph $G'$ with this property embeds $G$. We construct $2^{\aleph_0}$ path-minimal graphs whose ages are pairwise incomparable with set inclusion and which are wqo. Our construction is based on uniformly recurrent sequences and lexicographical sums of labelled graphs.

math.CO↗

arXiv2022

Metric properties of incomparability graphs with an emphasis on paths

We describe some metric properties of incomparability graphs. We consider the problem of the existence of infinite paths, either induced or isometric, in the incomparability graph of a poset. Among other things, we show that if the incomparability graph of a poset is connected and has infinite diameter, then it contains an infinite induced path. Furthermore, if the diameter of the set of vertices of degree at least $3$ is infinite, then the graph contains as an induced subgraph either a comb or a kite.

math.CO↗

arXiv2021

Hereditary rigidity, separation and density In memory of Professor I.G. Rosenberg

We continue the investigation of systems of hereditarily rigid relations started in Couceiro, Haddad, Pouzet and Schölzel [1]. We observe that on a set $V$ with $m$ elements, there is a hereditarily rigid set $\mathcal R$ made of $n$ tournaments if and only if $m(m-1)\leq 2^n$. We ask if the same inequality holds when the tournaments are replaced by linear orders. This problem has an equivalent formulation in terms of separation of linear orders. Let $h_{\rm Lin}(m)$ be the least cardinal $n$ such that there is a family $\mathcal R$ of $n$ linear orders on an $m$-element set $V$ such that any two distinct ordered pairs of distinct elements of $V$ are separated by some member of $\mathcal R$, then $ \lceil \log_2 (m(m-1))\rceil\leq h_{\rm Lin}(m)$ with equality if $m\leq 7$. We ask whether the equality holds for every $m$. We prove that $h_{\rm Lin}(m+1)\leq h_{\rm Lin}(m)+1$. If $V$ is infinite, we show that $h_{\rm Lin}(m)= \aleph_0$ for $m\leq 2^{\aleph_0}$. More generally, we prove that the two equalities $h_{\rm Lin}(m)= log_2 (m)= d({\rm Lin}(V))$ hold, where $\log_2 (m)$ is the least cardinal $μ$ such that $m\leq 2^μ$, and $d({\rm Lin}(V))$ is the topological density of the set ${\rm Lin}(V)$ of linear orders on $V$ (viewed as a subset of the power set $\mathcal{P}(V\times V)$ equipped with the product topology). These equalities follow from the {\it Generalized Continuum Hypothesis}, but we do not know whether they hold without any set theoretical hypothesis.

cs.DM↗

arXiv2020

Siblings of countable cographs

We show that every countable cograph has either one or infinitely many siblings. This answers, very partially, a conjecture of Thomassé. The main tools are the notion of well quasi ordering and the correspondence between cographs and some labelled ordered trees.

math.CO↗

arXiv2020

The poset of copies for automorphism groups of countable relational structures

Let $\mathrm{G}$ be a subgroup of the symmetric group $\mathfrak S(U)$ of all permutations of a countable set $U$. Let $\overline{\mathrm{G}}$ be the topological closure of $\mathrm{G}$ in the function topology on $U^U$. We initiate the study of the poset $\overline{\mathrm{G}}[U]:=\{f[U]\mid f\in \overline{\mathrm{G}}\}$ of images of the functions in $\overline{\mathrm{G}}$, being ordered under inclusion. This set $\overline{\mathrm{G}}[U]$ of subsets of the set $U$ will be called the \emph{poset of copies for} the group $\mathrm{G}$. A denomination being justified by the fact that for every subgroup $\mathrm{G}$ of the symmetric group $\mathfrak S(U)$ there exists a homogeneous relational structure $R$ on $U$ such that $\overline G$ is the set of embeddings of the homogeneous structure $R$ into itself and $\overline{\mathrm{G}}[U]$ is the set of copies of $R$ in $R$ and that the set of bijections $\overline G\cap \mathfrak S(U)$ of $U$ to $U$ forms the group of automorphisms of $\mathrm{R}$.

math.CO↗

arXiv2020

Generalized metric spaces. Relations with graphs, ordered sets and automata : A survey

In this survey we present a generalization of the notion of metric space and some applications to discrete structures as graphs, ordered sets and transition systems. Results in that direction started in the middle eighties based on the impulse given by Quilliot (1983). Graphs and ordered sets were considered as kind of metric spaces, where - instead of real numbers - the values of the distance functions $d$ belong to an ordered semigroup equipped with an involution. In this frame, maps preserving graphs or posets are exactly the nonexpansive mappings (that is the maps $f$ such that $d(f(x),f(y))\leq d(x,y)$, for all $x,y$). It was observed that many known results on retractions and fixed point property for classical metric spaces (whose morphisms are the nonexpansive mappings) are also valid for these spaces. For example, the characterization of absolute retracts, by Aronszajn and Panitchpakdi (1956), the construction of the injective envelope by Isbell (1965) and the fixed point theorem of Sine and Soardi (1979) translate into the Banaschewski-Bruns theorem (1967), the MacNeille completion of a poset (1933) and the famous Tarski fixed point theorem (1955). This prompted an analysis of several classes of discrete structures from a metric point of view. In this paper, we report the results obtained over the years with a particular emphasis on the fixed point property.

math.CO↗

arXiv2019

Injective envelopes of transition systems and Ferrers languages

We consider reflexive and involutive transition systems over an ordered alphabet $A$ equipped with an involution. We give a description of the injective envelope of any two-element set in terms of Galois lattice, from which we derive a test of its finiteness. Our description leads to the notion of Ferrers language.

math.CO↗

arXiv2019

Siblings of an $\aleph_0$-categorical relational structure

A sibling of a relational structure $R$ is any structure $S$ which can be embedded into $R$ and, vice versa, in which $R$ can be embedded. Let $sib(R)$ be the number of siblings of $R$, these siblings being counted up to isomorphism. Thomassé conjectured that for countable relational structures made of at most countably many relations, $sib(R)$ is either $1$, countably infinite, or the size of the continuum; but even showing the special case $sib(R)=1$ or infinite is unsettled when $R$ is a countable tree. This is related to Bonato-Tardif conjecture asserting that for every tree $T$ the number of trees which are sibling of $T$ is either one or infinite. We prove that if $R$ is countable and $\aleph_{0}$-categorical, then indeed $sib(R)$ is one or infinite. Furthermore, $sib(R)$ is one if and only if $R$ is finitely partitionable in the sense of Hodkinson and Macpherson. The key tools in our proof are the notion of monomorphic decomposition of a relational structure introduced in a paper by Pouzet and Thiéry 2013 and studied further by Oudrar and Pouzet 2015, and a result of Frasnay 1984.

math.LO↗