Finite approximation: for a Q̄-subspace Y ⊆ ℂⁿ and B ∈ ℕ, one finite-dimensional Y₀ ⊆ Y has dim Y₀/(Y₀ ∩ V) ≥ min(dim Y₁/(Y₁ ∩ V), B) for all ℂ-subspaces V and finite-dimensional Y₁ ⊆ Y
ProvedDiazModulus.fv_finite_approxdiaz-modulus-leannumber-theorytranscendence
Here is the field of algebraic numbers, the set of logarithms of algebraic numbers ( with algebraic), and the -vector space spanned by and . A subspace is rational if it is spanned by vectors with algebraic coordinates.
For a -subspace and a -subspace of , is a -subspace; dimensions are over for -subspaces and over for -subspaces. For a finite-dimensional -subspace and a -subspace write .
Let be a -subspace of any dimension and a natural number. Then there is a finite-dimensional -subspace such that
for every -subspace and every finite-dimensional -subspace . The same serves for every .
Preamble
import Definitions.Def_DiazModulus
Formal statement
open Complex ComplexConjugate
namespace DiazModulus
theorem fv_finite_approx {n : ℕ} (Y : Submodule (↥Qbar) (Fin n → ℂ)) (B : ℕ) :
∃ Y₀ : Submodule (↥Qbar) (Fin n → ℂ), Y₀ ≤ Y ∧ FiniteDimensional (↥Qbar) ↥Y₀ ∧
∀ (V : Submodule ℂ (Fin n → ℂ)) (Y₁ : Submodule (↥Qbar) (Fin n → ℂ)),
Y₁ ≤ Y → FiniteDimensional (↥Qbar) ↥Y₁ →
min (Module.finrank (↥Qbar) ↥Y₁ -
Module.finrank (↥Qbar) ↥(Y₁ ⊓ V.restrictScalars (↥Qbar))) B ≤
Module.finrank (↥Qbar) ↥Y₀ -
Module.finrank (↥Qbar) ↥(Y₀ ⊓ V.restrictScalars (↥Qbar)) := by
sorry
end DiazModulus
Source
Step of the proof of M. Waldschmidt, Further variations on the six exponentials theorem, Hardy-Ramanujan J. 28 (2005), 1–9, Th. 2.1 (p. 5), for every Q̄-subspace Y: one finite-dimensional Y₀ ⊆ Y that serves for every ℂ-subspace V at once, chosen by maximising a bounded natural number attached to the intersection of the ℂ-subspaces V with ρ(Z, V) < B; from DiazModulus.fv_rho_excess and DiazModulus.fv_rho_sInf. Novelty: none claimed; routine. Formal proof: Diaz modulus mission, 14 November 2026 (C. Perassi).