Syntactic aspects of hypergraph polytopes

TL;DR

This paper develops a unified inductive tree notation for faces of hypergraph polytopes, including associahedra and permutohedra, enabling clearer combinatorial descriptions and analysis of their structural properties.

Contribution

It introduces a general tree-based notation for hypergraph polytopes' faces, extending previous combinatorial tools and providing criteria to distinguish types of face truncations.

Findings

Unified notation for faces of various hypergraph polytopes

Criterion to identify edges from different associativity types

Alternative proofs of existing hypergraph polytope properties

Abstract

This paper introduces an inductively defined tree notation for all the faces of polytopes arising from a simplex by truncations. This notation allows us to view inclusion of faces as the process of contracting tree edges. Our notation instantiates to the well-known notations for the faces of associahedra and permutohedra. Various authors have independently introduced combinatorial tools for describing such polytopes. We build on the particular approach developed by Dosen and Petric, who used the formalism of hypergraphs to describe the interval of polytopes from the simplex to the permutohedron. This interval was further stretched by Petric to allow truncations of faces that are themselves obtained by truncations, and iteratively so. Our notation applies to all these polytopes. We illustrate this by showing that it instantiates to a notation for the faces of the permutohedron-based associahedra, that consists of parenthesised words with holes. Dosen and Petric have exhibited some families of hypergraph polytopes (associahedra, permutohedra, and hemiassociahedra) describing the coherences, and the coherences between coherences etc., arising by weakening sequential and parallel associativity of operadic composition. We complement their work with a criterion allowing us to recover the information whether edges of these "operadic polytopes" come from sequential, or from parallel associativity. We also give alternative proofs for some of the original results of Dosen and Petric.