Eq. (3.7) — a blossom inequality is violated iff
ProvedPadbergRao.UpperBound.eq_3_7combinatorial-optimizationgraph-theorymatchingp2o-batch-p200bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be a finite simple graph with node capacities and edge bounds , let and let be its node slacks. For every and every ,
where and is the sum of over the cut-set of .
The right-hand side is a cut-set condition: edges of appear with value , edges of in complemented form , and each node of contributes its slack. This rewriting of the violated blossom inequality is what the proof of Theorem 3.1 uses.
Formalization Note The equivalence is stated for an arbitrary real vector ; the oddness of plays no role in it.
Preamble
import Mathlib import Definitions.Def_PadbergRao_UpperBound_bMatchingSystem
Formal statement
namespace PadbergRao.UpperBound
/-- Padberg–Rao (1982), p. 75, Eq. (3.7): for `T ⊆ (W : V − W)`, the point `x̄` violates the
blossom inequality (3.3), `x̄(W) + x̄(T) > ½(b(W) + d(T) − 1)`, if and only if
`x̄(W : V − W) + d(T) − 2x̄(T) + s̄(W) < 1`. -/
theorem eq_3_7 {V : Type*} [Fintype V] [DecidableEq V] (G : SimpleGraph V) [DecidableRel G.Adj]
(b : V → ℕ) (d : Sym2 V → ℕ) (x : Sym2 V → ℝ) (W : Finset V) (T : Finset (Sym2 V))
(hT : T ⊆ cutEdges G W) :
((∑ i ∈ W, (b i : ℝ)) + (∑ e ∈ T, (d e : ℝ)) - 1) / 2
< (∑ e ∈ edgesWithin G W, x e) + ∑ e ∈ T, x e ↔
(∑ e ∈ cutEdges G W, x e) + (∑ e ∈ T, (d e : ℝ)) - 2 * (∑ e ∈ T, x e)
+ (∑ i ∈ W, slack G b x i) < 1 := by sorry
end PadbergRao.UpperBound
Source
Padberg, Rao, Odd Minimum Cut-Sets and b-Matchings, Math. Oper. Res. 7 (1982), p. 75, Section 3, Eq. (3.7) (equivalence with (3.3) used on p. 77 in the proof of Theorem 3.1)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.