Semantic Limits of Dense Combinatorial Objects

TL;DR

This paper develops a general theory of dense limit objects for combinatorial structures using model theory, unifying various existing concepts like graphons and hypergraphons, and extends fundamental properties such as existence and uniqueness.

Contribution

It introduces a model-theoretic framework for dense combinatorial limits that unifies and generalizes previous approaches, including graphons and hypergraphons.

Findings

Unified framework for dense combinatorial limits

Extended existence and uniqueness properties

Provided a proof of the continuous Induced Removal Lemma

Abstract

The theory of limits of discrete combinatorial objects has been thriving for the last decade or so. The syntactic, algebraic approach to the subject is popularly known as "flag algebras", while the semantic, geometric one is often associated with the name ``graph limits''. The language of graph limits is generally more intuitive and expressible, but a price that one has to pay for it is that it is better suited for the case of ordinary graphs than for more general combinatorial objects. Accordingly, there have been several attempts in the literature, of varying degree of generality, to define limit objects for more complicated combinatorial structures. This paper is another attempt at a workable general theory of dense limit objects. Unlike previous efforts in this direction (with notable exception of [Ashwini Aroskar and James Cummings. Limits, regularity and removal for finite structures. Technical Report arXiv:1412.2014 [math.LO], arXiv e-print, 2014.]), we base our account on the same concepts from the first-order logic and the model theory as in the theory of flag algebras. We show how our definition naturally encompasses a host of previously considered cases (graphons, hypergraphons, digraphons, permutons, posetons, colored graphs, etc.), and we extend the fundamental properties of existence and uniqueness to this more general case. We also give an intuitive general proof of the continuous version of the Induced Removal Lemma based on the completeness theorem for propositional calculus. We capitalize on the notion of an open interpretation that often allows to transfer methods and results from one situation to another. Again, we show that some previous arguments can be quite naturally framed using this language.