Degrees of interior polynomials and parking function enumerators
Tamás Kálmán, Lilla Tóthmérész
TL;DR
The work establishes precise degree formulas for interior polynomials of digraphs and their matroid generalizations, showing $\deg I_G = |V|-1-\nu(G)$ where $\nu(G)$ is the minimum cardinality of a dijoin, and that the leading coefficient counts net degree vectors of minimum dijoins. By duality, these results translate to parking-function enumerators and greedoid polynomials, yielding $\deg$ of the parking-function enumerator for an Eulerian digraph as $|E|-|V|+1-\minfas(G)$ and extending to arbitrary rooted digraphs via $\mathrm{minfas}(G,r)$. The paper further generalizes to oriented regular matroids, giving analogous degree formulas and a facet description of extended root polytopes, thereby unifying graph- and matroid-theoretic perspectives. The connections illuminate how combinatorial optimizations (dijoins, cuts, and feedback arc sets) govern polynomial degrees in Ehrhart-theoretic encodings, with implications for counting generalized parking functions and analyzing greedoid polynomials across broad combinatorial structures.
Abstract
The interior polynomial of a directed graph is defined as the $h^*$-polynomial of the graph's (extended) root polytope, and it displays several attractive properties. Here we express its degree in terms of the minimum cardinality of a directed join, and give a formula for the leading coefficient. We present natural generalizations of these results to oriented regular matroids; in the process we also give a facet description for the extended root polytope of an oriented regular matroid. By duality, our expression for the degree of the interior polynomial implies a formula for the degree of the parking function enumerator of an Eulerian directed graph (which is equivalent to the greedoid polynomial of the corresponding branching greedoid). We extend that result to obtain the degree of the parking function enumerator of an arbitrary rooted directed graph in terms of the minimum cardinality of a certain type of feedback arc set.
