This repository formalizes a solution to
Erdős Problem #1196, which asks whether every primitive set
supported on [x, ∞) satisfies
The formalized result proves the quantitative estimate
for every primitive set A ⊆ ℕ ∩ [x, ∞). The argument follows the explicit sub-Markov-chain
construction in source.tex, where the visiting probabilities are
proportional to 1 / (n log n).
PrimitiveSetsAboveX.mainTheoremin PrimitiveSetsAboveX/Main.lean formalizes the main bound for Erdős Problem #1196.Erdos1196.erdos_1196in PrimitiveSetsAboveX/FormalConjecturesErdos1196.lean packages the result in the same mathematical form as theformal-conjecturesstatement of Erdős Problem #1196.- The repository includes the Lean development, a blueprint, and the TeX source for the writeup.
- PrimitiveSetsAboveX/Basic.lean and PrimitiveSetsAboveX/PrimitiveWeight.lean define primitive sets, weights, and the weighted series.
- PrimitiveSetsAboveX/Preliminaries.lean, PrimitiveSetsAboveX/NormalizationCore.lean, and PrimitiveSetsAboveX/NormalizationSmallPrime.lean establish the analytic normalization estimates.
- PrimitiveSetsAboveX/Markov.lean develops the sub-Markov chain and its explicit visit formula.
- PrimitiveSetsAboveX/HitMass.lean implement the first-hit mass argument leading to the final weighted-series bound.
- PrimitiveSetsAboveX/Main.lean assembles the full theorem.
- PrimitiveSetsAboveX/FormalConjecturesErdos1196.lean
provides a bridge theorem matching the
formal-conjecturesstatement for Erdős Problem #1196. - blueprint/src/content.tex and blueprint/src/sections/ contain the blueprint source.
- source.tex contains the TeX source for the accompanying writeup.
- Erdős Problem #1196
- Main theorem in Lean
- Formal-conjectures style theorem in Lean
- TeX writeup source
- Blueprint source
Compile the Lean development:
lake exe cache get && lake buildBuild the blueprint PDF:
uvx leanblueprint pdfBuild and serve the blueprint website:
uvx leanblueprint web && uvx leanblueprint serveBuild the TeX writeup PDF:
latexmk -pdf source.tex