English

Counting Dope Matrices

Combinatorics 2022-12-13 v2

Abstract

For a polynomial PP of degree nn and an mm-tuple Λ=(λ1,,λm)\Lambda=(\lambda_1,\dots,\lambda_m) of distinct complex numbers, the dope matrix of PP with respect to Λ\Lambda is DP(Λ)=(δij)i[1,m],j[0,n]D_P(\Lambda)=(\delta_{ij})_{i\in [1,m],j\in[0,n]}, where δij=1\delta_{ij}=1 if P(j)(λi)=0P^{(j)}(\lambda_i)=0, and δij=0\delta_{ij}=0 otherwise. Our first result is a combinatorial characterization of the 22-row dope matrices (for all pairs Λ\Lambda); using this characterization, we solve the associated enumeration problem. We also give upper bounds on the number of m×(n+1)m\times(n+1) dope matrices, and we show that the number of m×(n+1)m \times (n+1) dope matrices for a fixed mm-tuple Λ\Lambda is maximized when Λ\Lambda is generic. Finally, we resolve an ``extension'' problem of Nathanson and present several open problems.

Keywords

Cite

@article{arxiv.2205.09302,
  title  = {Counting Dope Matrices},
  author = {Noga Alon and Noah Kravitz and Kevin O'Bryant},
  journal= {arXiv preprint arXiv:2205.09302},
  year   = {2022}
}