Termwise cancellation for products of naturals bounded by positive factors.
ProvedRound10.prod_eq_prod_of_leaether-cataloggeometry
Termwise cancellation for products of naturals bounded by positive factors.
theorem Round10.prod_eq_prod_of_le: ∀ (P : Finset ℕ) (f g : ℕ → ℕ), (∀ i ∈ P, f i ≤ g i) →
(∀ i ∈ P, 0 < g i) → ∏ i ∈ P, f i = ∏ i ∈ P, g i → ∀ i ∈ P, f i = g i := by sorry
Formalization Note Transplanted verbatim from the Aether Catalog source Geometry/Round10Closures/SquarefreeTrace.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.
Preamble
-- Thm stub generated from Geometry/Round10Closures/SquarefreeTrace.lean
import Mathlib
/-
Round-10 Closures — Part VIII (cycle 3): the trace lemma beyond semiprimes.
Cycle 3 pushes the classification off the semiprime case: the free-witness family is
multiplicative, so for every squarefree modulus `N = ∏_{r ∈ P} r` (a finite set of distinct
primes) the witness is the product of the local gcd-residue coordinates,
R_k(N) = ∏_{r ∈ P} gcd(k, r - 1).
Two consequences are recorded:
* the population of square roots of unity is `2^ω(N)` for odd squarefree `N`, so the
residue coordinate *counts the prime factors* — the classified coordinate already knows
`ω(N)`, while it still cannot name a single factor without aggregation;
* the witness of the exponent `k` is still bounded by `k^{ω(N)}`, so the bounded-exponent
barrier of `JointClosure.lean` degrades only polynomially in the number of factors.
-/
/-! ### The Carmichael threshold for squarefree moduli (cycle 4)
The completeness analysis of `AggregationCost.lean` generalises verbatim: a free witness of
a squarefree modulus is maximal exactly at the multiples of `lcm_{r ∈ P} (r-1)`, the
Carmichael exponent of `N`. So the aggregation depth of the classical channel is the
Carmichael function, for every squarefree modulus and not just for semiprimes. -/Formal statement
theorem Round10.prod_eq_prod_of_le: ∀ (P : Finset ℕ) (f g : ℕ → ℕ), (∀ i ∈ P, f i ≤ g i) →
(∀ i ∈ P, 0 < g i) → ∏ i ∈ P, f i = ∏ i ∈ P, g i → ∀ i ∈ P, f i = g i := by sorrySource