Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedjudge pass (gpt-6-luna)
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.

Dirichlet parameters add under coarsening

Statement

Assume AC. Let P=(P1,…,Pk) have the Dirichlet distribution with positive parameters (a1,…,ak). If I1,…,Ir is a partition of {1,…,k} into nonempty groups, then (∑j∈I1Pj,…,∑j∈IrPj)∼Dirichlet⁡(b1,…,br),bℓ=∑j∈Iℓaj. When r=1, the grouped vector is the point mass at 1.

Facts & Assumptions

Given: Full AC; positive a1,…,ak; and a partition of the coordinate indices into nonempty groups.

[A1]

Full AC is assumed. The Axiom of Choice.

[A2]

Full AC implies the Countable Choice required by Gamma density normalization. AC implies DC implies countable choice The Axiom of Countable Choice (ACω).

[F1]

On a finite discrete standard Borel space, the normalized-gamma construction realizes a positive-parameter Dirichlet vector as independent Gamma⁡(aj,1) increments divided by their sum. Existence of a Dirichlet process by normalized gamma Poisson random measure.

[F2]

Under Countable Choice, a shape-rate Gamma(a,1) variable has density za−1e−z/Γ(a), normalized for a>0; the Euler Gamma value is finite and positive. Shape-rate Gamma distribution The real Gamma function by Euler's integral Euler's Gamma integral converges exactly for positive real parameters.

[F3]

Under Countable Choice, the nonnegative Lebesgue change-of-variables formula holds for C1 diffeomorphisms. A C^1 diffeomorphism satisfies the change-of-variables formula for nonnegative Lebesgue measurable functions.

[F4]

Finite measures with an exponential moment and the same Laplace transform on a neighborhood of zero are equal under AC. Uniqueness of a multivariate exponential transform.

[F5]

Finite products of probability measures exist and satisfy the rectangle product formula. For sigma-finite factors, the product measure exists, has the rectangle formula, is sigma-finite, and is unique.

[F6]

The finite-dimensional Dirichlet law is the positive-parameter simplex law specified in Dirichlet distribution, including the one-category case.

[F7]

Every nonempty finite discrete space is standard Borel. Dirac set functions are probability measures; finite nonnegative weighted sums of them are measures. A countable discrete measurable space is standard Borel A Dirac set function is a probability measure Nonnegative scalar multiples and countable weighted sums of measures are measures.

Proof

Given: The hypotheses in the Statement.

1.1A1F1F5F6F7givenconstruct

The nonempty finite discrete space S={1,…,k} is standard Borel by [F7]. Define α(B)=∑j∈Baj for B⊆S, equivalently α=∑j=1kajδj. By [F7] this is a measure, with finite positive total ∑j=1kaj, so [F1] applies to (S,P(S),α). Thus the law of P is that of Zj/∑h=1kZh, where the Zj are independent Gamma⁡(aj,1) variables, and it is the Dirichlet law [F6].

2.1A1A2F2F3F5step 1.1algebra

For u<1, [F2] and the change of variables t=(1−u)z in [F3] give EeuZj=Γ(aj)−1∫0∞zaj−1e−(1−u)zdz=(1−u)−aj. For Wℓ=∑j∈IℓZj, independence therefore gives the joint moment-generating function ∏ℓ=1r(1−uℓ)−bℓ for ∣uℓ∣<δ<1, since the groups use disjoint coordinates in the finite product law [F5].

3.1A1A2F4F5step 2.1

By [F5], the product law of independent Gamma⁡(bℓ,1) variables exists and has the same joint transform. Both laws have finite exponential moment at every 0<δ<1, since the transform at (δ,…,δ) is finite. Applying [F4] to the difference of these two finite measures shows that (W1,…,Wr) has that product Gamma law.

4.1F1step 1.1step 3.1algebracases∎

Normalize the grouped variables by their common positive total. The resulting vector is Dirichlet with parameters (b1,…,br) by [F1], while Wℓ/∑hWh=∑j∈IℓPj by step 1.1. This proves the claim for r≥2. For r=1 the only coordinate sum is ∑j=1kPj=1, which is the stated one-category law.

Depends on

Used by

Dependency tree · two levels

105 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources