The Tree-Forest Ratio

05/23/2022
by   Oliver Knill, et al.
0

The number of rooted spanning forests divided by the number of spanning rooted trees in a graph G with Kirchhoff matrix K is the spectral quantity tau(G)= det(1+K)/det(K) of G by the matrix tree and matrix forest theorems. We prove that that under Barycentric refinements, the tree index T(G)=log(det(K))/n and forest index F(G)=log(det(1+K))/n and so the tree-forest index i=F-G=log(tau(G))/n converge to numbers that only depend on the size of the maximal clique in the graph. In the 1-dimensional case, all numbers are known: T(G)=0, F(G)=i(G) =2 log(phi), where phi is the golden ratio. The convergent proof uses the Barycentral limit theorem assuring the Kirchhoff spectrum converges weakly to a measure dk on the positive real axis that only depends on dimension of G. Trees and forests indices are potential values i = U(-1)-U(0) for the subharmonic function U(z)=int_R log|x-z| dk(x) defined by the Riesz measure dk=Delta U which only depends on the dimension of G. The potential U(z) is defined for all z away from the support of dk and finite at z=0. Convergence follows from the tail estimate k[x,infty] < C exp(-a x) where the decay rate a only depends on the maximal dimension. With the normalized zeta function zeta(s) = (1/n) sum_k lambda_k^-s, we have for all finite graphs of maximal dimension larger than 1 the identity i(G) = sum_t (-1)^(s+1) zeta(s)/s. The limiting zeta function zeta(s) = int_R x^(-s) dk(x) is analytic in s for s<0. The Hurwitz spectral zeta function zeta_z(s)=U_s(z) = int_R (x-z)^(-s) dk(x) complements U(z) = int_R log(x-z) dk(x) and is analytic for z in C - R^+ and for fixed z in C-R^+ is an entire function in s in C.

READ FULL TEXT

page 3

page 18

research
10/19/2022

Subtractive random forests

Motivated by online recommendation systems, we study a family of random ...
research
01/04/2013

The Sum-over-Forests density index: identifying dense regions in a graph

This work introduces a novel nonparametric density index defined on grap...
research
08/09/2005

Matrices of Forests and the Analysis of Digraphs

The matrices of spanning rooted forests are studied as a tool for analys...
research
02/25/2021

Spanning Tree Constrained Determinantal Point Processes are Hard to (Approximately) Evaluate

We consider determinantal point processes (DPPs) constrained by spanning...
research
03/22/2015

Adaptive Concentration of Regression Trees, with Application to Random Forests

We study the convergence of the predictive surface of regression trees a...
research
01/18/2021

Graph complements of circular graphs

Graph complements G(n) of cyclic graphs are circulant, vertex-transitive...
research
09/10/2019

Local Large Deviation Principle, Large Deviation Principle for the Signal -to- Interference and Noise Ratio Graph Models

Given devices space D, an intensity measure λ∈(0,∞), transition kernel f...

Please sign up or login with your details

Forgot password? Click here to reset