Paperlane

no. 14 · Friday 13 November 2026

Generalized Discrete Fractal Dimensions of Graph Families

Jiantao Liu, Cheng Zeng, Yumei Xue and Qingcheng Zeng

math.CO · v2

Abstract

We study generalized discrete fractal dimensions of sequences of finite connected graphs whose diameters tend to infinity. Using normalized shortest-path metrics and a general scale function, we define regularized energy dimensions for vertex probability measures and their optimized versions. We then establish a variational principle that identifies the optimized energy dimension with the supremum of ball-mass concentration dimensions. We also establish monotonicity in the moment order, stability under uniformly comparable measures, and upper bounds in terms of Rényi entropy growth and the upper mass dimension. Cluster estimates and explicit graph constructions show that local vertex concentration can lower the uniform energy dimension without changing the mass dimension. Finally, for uniformly Ahlfors-regular graph families, we obtain an integral criterion for energy boundedness and explicit formulas for the critical dimensions. These results clarify how global growth, local distance distributions, and the choice of scale function jointly determine discrete fractal dimensions.

fields
Combinatorics, Metric Geometry
pdf ↗ page arXiv:2610.39187 ↗