Answers are grounded in this paper and cite their sources.
Develops a proof connecting transversal matroids and Hall's theorem, enhancing computational geometry in mathematics software.
Mathlib has Hall's marriage theorem. Mathlib has matroids. In the revision this work was checked against (84fdd0bf, 2026-07-15) nothing joins the two: every matroid in Matroid/Constructions.lean is trivial, and none is built from combinatorial data. This paper closes that gap for the classical bridge between them, the Transversal Matroid Theorem of Edmonds–Fulkerson and, independently, Mirsky–Perfect: for a finite family of finite sets, the partial transversals are exactly the independent sets of a matroid. The contribution is not that Lean accepts the proof; it is the shape of the proof. Rather than certify an augmenting-path algorithm, the development runs Hall's obstruction method and isolates one object in its own layer: a tight set R, one whose compatible slots are exactly saturated, |N(R)| = |R|. That factoring pays a dividend the classical statement does not advertise. Tight sets are closed under intersection as well as union, hence form a lattice, so every partial transversal carries a single largest tight subset maxTight A I, and one containment test against it decides every insertion. Put in Mathlib's own vocabulary, that same set computes the matroid's closure operator on independent sets. 645 lines of Lean, 34 theorems, no sorry, admit or native_decide; #print axioms on all six headline theorems reports only propext, Classical.choice, Quot.sound. English and Spanish editions, plus a reproduction bundle with the Lean sources, the generated axiom report, an exhaustive SageMath cross-check and the deterministic figure generator. The target was drawn from Theorem of the Day's list of theorems by women mathematicians; Hazel Perfect, whose name puts it there, went on to introduce the gammoids — the class containing the transversal matroids — the year before her doctorate, Studies in Transversal Theory. Priority claims are dated, itemised and falsifiable in the paper's own priority section: they are audit statuses, not proofs of absence.
No takes yet. Share an insight, caveat, or question.
Carmen Muñoz (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: