A tight difference-constraint path must saturate a target inequality saturated by another feasible path
ProvedHirsch.network_tight_path_forces_target_tightdifference-constraintshirsch-conjecturenetwork-routingpolyhedra
Suppose u saturates every consecutive difference constraint along a directed path, v satisfies all the same path constraints, v saturates a shortcut/target inequality with bound C, and u satisfies that target inequality. Then u must also saturate the target inequality. This is the scalar core used in the network-potential routing argument to force a backward tight row during target insertion.
Preamble
import Mathlib set_option autoImplicit false noncomputable section
Formal statement
theorem Hirsch.network_tight_path_forces_target_tight
(n : ℕ) (u v : ℕ → ℝ) (cost : Fin n → ℝ) (C : ℝ)
(hu : ∀ i : Fin n, u (i.val + 1) - u i.val = cost i)
(hv : ∀ i : Fin n, v (i.val + 1) - v i.val ≤ cost i)
(hvtarget : v n - v 0 = C)
(hutarget : u n - u 0 ≤ C) :
u n - u 0 = C := by sorrySource
Standalone Mathlib-only restatement of the target-tightness primitive developed in PR #210.