Equation (2) has a unique root
ProvedConstrNestedLogit.Stitch.equation_2_uniqueFix a finite collection of candidate assortments in each nest, with at least one candidate per nest. Let be the total preference weight of the products in , their conditional expected revenue, the nest dissimilarity, and the weight of no purchase. Then there is exactly one real number satisfying equation (2):
This identifies the scalar value used to combine the candidate assortments. The paper states existence and uniqueness in the paragraph following Lemma 1 without a separate proof.
Formalization Note Nests form a finite type, and each is a nonempty finite set of assortments. Products are indexed from zero. The main model uses ; strictly positive product weights and make the probability model well defined. Revenues have no ordering or sign restriction.
import Mathlib import Definitions.Def_NestedLogitVariants_LP_Model
namespace ConstrNestedLogit.Stitch
/-- Section 2, p. 12: equation (2) has a unique solution for nonempty finite candidate
collections. -/
theorem equation_2_unique {ι : Type*} [Fintype ι] [DecidableEq ι] {n : ℕ}
(I : NestedLogitVariants.LP.Instance ι n)
(A : ι → Finset (Finset (Fin n)))
(hA : ∀ i, (A i).Nonempty)
(hvnp : ∀ i, I.vnp i = 0)
(hv0 : 0 < I.v0)
(hv : ∀ i j, 0 < I.v i j)
(hγ : ∀ i, 0 < I.γ i ∧ I.γ i ≤ 1) :
∃! z : ℝ, I.v0 * z = ∑ i, (A i).sup' (hA i) (fun S =>
NestedLogitVariants.LP.nestWeight I i S * (NestedLogitVariants.LP.R I i S - z)) := by sorry
end ConstrNestedLogit.Stitch
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.