paper

Cluster-Graph Edit Distance: Optimal Explicit Embeddings, Metric Proxies, and Complexity

arXiv:2608.17990

Abstract

The cluster graphs on vertices, the disjoint unions of complete graphs, have the integer partitions of as their isomorphism classes, and the quotient edit distance makes that set a metric space. Its geometry and its complexity both issue from one identity: is an affine function of the maximum of over the contingency tables with margins and . Our main result is an explicit optimal embedding. The weighted dyadic sums of the Ferrers staircase, taken at the critical exponent , give a map into that acts on a single partition and is computable in time, and its distortion is . That order is optimal, since : the lower half follows from a -dimensional Hamming cube of partitions and Enflo's theorem, so the determination needs no other external input. The analytic core is a scale-free inverse inequality for every integer sequence with and : its critical dyadic energy is at least . Combinatorially the same identity yields two explicit models, the vertex-mass metric on sorted degree sequences with and the block-energy metric with , both constants optimal; hence , and an -time algorithm returns an alignment of cost below carrying the certificate . Computationally, deciding is strongly NP-complete and admits no FPTAS, while the farthest alignment is polynomial-time solvable. The best constant in the inverse inequality remains open; an exactly solvable chirp family caps it at .

49 pages, 6 figures. v2: the critical inverse-energy conjecture of v1 is now proved, as a scale-free inverse theorem on the full cone of closed quantized integer sequences (Theorem 6.2)