Erdős problem 346
Let be a set of integers such that is complete for any finite subset and not complete for any infinite subset . If for all , must ? Under the reading where the ratio limit is assumed to exist, a Lean-verified argument forces the limit to be the golden ratio; a separate construction disproves the literal statement where convergence is not assumed.
Sources
FormalConjectures/ErdosProblems/
346.lean
Retained formal statement
Erdős and Graham [ErGr80] also say that it is not hard to construct very irregular sequences satisfying the aforementioned properties.
∃ A, IsAddStronglyCompleteNatSeq A ∧ (∀ B ⊆ Set.range A, B.Infinite → ¬IsAddComplete (Set.range A \ B)) ∧ Filter.liminf (fun n => ↑(A (n + 1)) / 2) Filter.atTop = 1 ∧ Filter.limsup (fun n => ↑(A (n + 1)) / ↑(A n)) Filter.atTop = ⊤SolvedStatement only, no proof