Edge tensor generators and vertex-incidence vectors of a graph
DefinitionAsymSpecConn_Supplement_graphGeneratoralgebraic-complexitygraph-tensorsp2o-gran-per-chapterp2o-plan-paperp2o-v1tensor-rank
Let be a finite simple graph on vertices . For each edge , its edge tensor generator is the two-dimensional unit tensor on legs and , with one-dimensional factors on every other leg. For a vertex , its incidence vector has coordinate precisely on edges incident with .
The ordered endpoint representation indexes each unoriented edge exactly once. Isolated vertices remain present and have incidence vector zero.
These are the graph tensors and convex-hull centers in Definition 4.11 and Theorem 4.12.
Definition code
import Mathlib
import Definitions.Def_GraphTensorCP_wt
import Definitions.Def_mme_tensor_quotient
set_option autoImplicit false
open scoped NNReal
namespace AsymSpecConn.Supplement
universe u
/-- The unoriented edges of a finite simple graph, represented by their uniquely ordered endpoints. -/
abbrev GraphEdge {k : ℕ} (H : SimpleGraph (Fin k)) :=
{e : Fin k × Fin k // e.1 < e.2 ∧ H.Adj e.1 e.2}
/-- The finite edge index type, obtained classically so theorem statements do not carry a
computability hypothesis on graph adjacency. -/
noncomputable instance graphEdgeFintype {k : ℕ} (H : SimpleGraph (Fin k)) :
Fintype (GraphEdge H) := Fintype.ofFinite _
/-- The generator `b_e` of Definition 4.11 (p.36): the two-dimensional unit tensor on edge
`e`, with one-dimensional factors on all other legs. This is the published edge tensor
`GraphTensorCP.edgeTensor`. -/
noncomputable def graphGenerator (K : Type u) [Field K] {k : ℕ} (H : SimpleGraph (Fin k))
(e : GraphEdge H) : MME.TensorQ K k :=
MME.TensorQ.toQ (GraphTensorCP.edgeTensor K e.1.1 e.1.2 2)
/-- The vertex-incidence vector `chi_w ∈ {0,1}^E` of p.36. Isolated vertices give the zero
vector; they are deliberately retained in Theorem 4.12. -/
def incidenceVector {k : ℕ} (H : SimpleGraph (Fin k)) (w : Fin k) : GraphEdge H → ℝ :=
fun e => if w = e.1.1 ∨ w = e.1.2 then 1 else 0
end AsymSpecConn.Supplement
Source
Vrana and Zuiddam, The asymptotic spectrum of tensors is connected, preprint (STOC 2027 submission, version of November 2, 2026, sha256 85a0b2ac698c902a02f2df11dd0671a5d9ca373aed822ae44b7e2640fc152c7d), Definition 4.11 and Theorem 4.12 p.36