/-
Copyright 2025 The Formal Conjectures Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
https://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
-/
module
public import FormalConjecturesUtilErdős Problem 477
References:
Sek59 Milan Sekanina, Замечания к фактoризации беcкoнечнoй цикличеcкoй группы, Czechoslovak Mathematical Journal, Vol. 9 (1959), No. 4, 485–495
The resolution is recorded at erdosproblems.com/477, with a proof exposition by T. F. Bloom of a construction found independently by several provers.
@[expose] public sectionopen Polynomial Setnamespace Erdos477set_option linter.style.formal_proof_palomar false inIs there a polynomial $f:\mathbb{Z}\to \mathbb{Z}$ of degree at least $2$ and a set $A\subset \mathbb{Z}$ such that for any $n\in \mathbb{Z}$ there is exactly one $a\in A$ and $b\in { f(k) : k\in\mathbb{Z}}$ such that $n=a+b$?
The answer is yes, contrary to the expectation of Erdős and Graham: such an A exists whenever
$f(n) = n^d$ for even $d \ge 6$.
The linked formal proof (Codex, following Price's exposition) exhibits a complement of
${k^6 : k \in \mathbb{Z}}$; it states uniqueness as ∃! p : ℤ × ℤ, p.1 ∈ A ∧ p.2 ∈ B ∧ p.1 + p.2 = n
and the degree condition as 2 ≤ f.natDegree.
@[category research solved, question, AMS 12, formal_proof using lean4 at
"https://github.com/plby/lean-proofs/blob/8822f7ddef30fadbd92e1c6ab4ed897af356af5e/src/latest/ErdosProblems/Erdos477.lean#L54"]
theorem erdos_477 :
∃ f : ℤ[X], 2 ≤ f.degree ∧ ∃ A : Set ℤ,
∀ z, ∃! ab ∈ A ×ˢ (Set.range f.eval), z = ab.1 + ab.2 := ⊢ ∃ f, 2 ≤ f.degree ∧ ∃ A, ∀ (z : ℤ), ∃! ab, (ab ∈ A ×ˢ range fun x ↦ eval x f) ∧ z = ab.1 + ab.2
All goals completed! 🐙There is no such $A$ for the polynomial $f(x) = X^2$.
This is shown in [Sek59].
@[category research solved, AMS 12]
theorem erdos_477.variants.S_sq :
letI f := X ^ 2
∀ A : Set ℤ, ∃ z, ¬ ∃! a ∈ A ×ˢ (Set.range f.eval), z = a.1 + a.2 := ⊢ ∀ (A : Set ℤ), ∃ z, ¬∃! a, (a ∈ A ×ˢ range fun x ↦ eval x (X ^ 2)) ∧ z = a.1 + a.2
All goals completed! 🐙There is no such $A$ for any polynomial $f(x) = aX^2 + bX + c$, if $a | b$ with $a \ne 0$ and $b \ne 0$. This was found be AlphaProof for the specific instance $X^2 - X + 1$ and then generalised.
@[category research solved, AMS 12]
theorem erdos_477.variants.degree_two_dvd_condition_b_ne_zero {a b c : ℤ} (ha : a ≠ 0) (hb : b ≠ 0)
(hab : a ∣ b) :
let f := a • X ^ 2 + b • X + C c
∀ A : Set ℤ, ∃ z, ¬ ∃! a ∈ A ×ˢ (Set.range f.eval), z = a.1 + a.2 := a:ℤb:ℤc:ℤha:a ≠ 0hb:b ≠ 0hab:a ∣ b⊢ let f := a • X ^ 2 + b • X + C c;
∀ (A : Set ℤ), ∃ z, ¬∃! a, (a ∈ A ×ˢ range fun x ↦ eval x f) ∧ z = a.1 + a.2
All goals completed! 🐙Probably there is no such $A$ for the polynomial $X^3$.
@[category research open, AMS 12]
theorem erdos_477.variants.X_pow_three :
letI f := X ^ 3
∀ A : Set ℤ, ∃ z, ¬ ∃! a ∈ A ×ˢ (Set.range f.eval), z = a.1 + a.2 := ⊢ ∀ (A : Set ℤ), ∃ z, ¬∃! a, (a ∈ A ×ˢ range fun x ↦ eval x (X ^ 3)) ∧ z = a.1 + a.2
All goals completed! 🐙set_option linter.style.formal_proof_palomar false inSekanina [Sek59] asked whether there is no such $A$ for $X^k$, for every $k \ge 2$. This is false: a complement exists for every even $k \ge 6$. The linked formal proof gives the case $k = 6$.
@[category research solved, question, AMS 12, formal_proof using lean4 at
"https://github.com/plby/lean-proofs/blob/8822f7ddef30fadbd92e1c6ab4ed897af356af5e/src/latest/ErdosProblems/Erdos477.lean#L42"]
theorem erdos_477.variants.monomial :
¬ ∀ (k : ℕ), 2 ≤ k →
letI f := X ^ k
∀ A : Set ℤ, ∃ z, ¬ ∃! a ∈ A ×ˢ (Set.range f.eval), z = a.1 + a.2 := ⊢ ¬∀ (k : ℕ), 2 ≤ k → ∀ (A : Set ℤ), ∃ z, ¬∃! a, (a ∈ A ×ˢ range fun x ↦ eval x (X ^ k)) ∧ z = a.1 + a.2
All goals completed! 🐙end Erdos477