How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
Supporting hyperplanes at relative boundary points
Statement
For nonempty , define . An affine Euclidean space is a translate of a finite-dimensional real linear subspace with the induced Euclidean metric. A supporting hyperplane at is for a nonzero vector , with in the set and all its points in one closed half-space. Let be a nonempty convex set in a finite-dimensional affine Euclidean space of dimension , let , and put . If (that is, and every relative neighborhood of meets ), then there is a nonzero linear functional on the translation space of such that for all . If has affine dimension at most , then every has at least two distinct ambient supporting hyperplanes containing .
Facts & Assumptions
Given: The stated finite-dimensional affine Euclidean space and convex set.
Heine-Borel in : with the Euclidean metric a subset of is compact if and only if it is closed and bounded, and the proof by bisection uses no choice principle; the same holds on the real line: closed bounded subsets of finite-dimensional Euclidean space are compact.
The Euclidean inner product on : Euclidean coordinates identify linear functionals with vectors and give the usual unit sphere.
A continuous real-valued function on a nonempty compact metric space is bounded and attains a greatest and a least value: a continuous real function on a nonempty compact metric space attains its minimum.
Unit spheres in finite-dimensional normed spaces are sequentially compact: unit spheres in finite-dimensional normed spaces are sequentially compact without a choice principle.
The rationals embed densely in the reals: rational coordinates are dense in real coordinates.
Reciprocals of the positive integers decrease to zero supplies the positive real sequence , , decreasing to zero with the canonical-natural field convention.
Proof
Fix affine coordinates . Closure preserves convexity without choosing approximating sequences: for and any , choose two points of within of them; their convex combination approximates within , so this point lies in . Enumerate rational -tuples canonically by increasing sums of absolute numerators and positive denominators, breaking the finite ties lexicographically. Density [F6] and the open complement of show that for each at least one enumerated point lies in . Let be the first such point. This is a uniquely specified sequence with , not a countable selection of arbitrary witnesses.
The nonempty compact set contains . By [F2] and [F4], squared distance to attains a minimum there, and every point outside that ball is farther than the candidate . Thus a nearest point exists. It is unique: if two distinct minimizers had squared distance , their midpoint would have squared distance , contradicting minimality and convexity. Consequently is also canonically determined. Since , put . Expanding the distance at and letting gives for every . Also .
Every has norm one. The exact choice-free sequential-compactness supplier [F5] gives a subsequence converging to a unit vector . This introduces no additional choice hypothesis.
For every , pass to the limit in along that subsequence. Continuity of the Euclidean pairing and give . The unit vector defines a nonzero supporting functional for , hence for , at .
If has affine hull of dimension at most , write its direction space as the span of independent vectors . Its orthogonal complement is the solution space of the homogeneous equations . Row reduction has at most pivot variables, leaving at least free variables. Assigning one free variable the value one and the others zero gives two independent solutions (their free-coordinate vectors are different unit vectors). Thus choose two nonparallel nonzero normals orthogonal to the direction space of . For each , the hyperplane contains and therefore supports at . The two hyperplanes are distinct because their normals are not parallel.
Depends on
Reciprocals of the positive integers decrease to zero
Unit spheres in finite-dimensional normed spaces are sequentially compact
A continuous real-valued function on a nonempty compact metric space is bounded and attains a greatest and a least value
The rationals embed densely in the reals
A convex subset of contains every line segment between two of its points
Heine-Borel in : with the Euclidean metric a subset of is compact if and only if it is closed and bounded, and the proof by bisection uses no choice principle; the same holds on the real line
The Euclidean inner product on