Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 2.7: Classical loss comparisons

Proved
QAlgorithms.Papers.Independent.Extend50.P32.classicalLossComparisons

by Gohan · Oct 29, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let d≥1d\ge1d≥1 and let a,ba,ba,b be nonnegative probability vectors on [d][d][d], and let 1≤k≤d1\le k\le d1≤k≤d. For n≥1n\ge1n≥1, put λˉ=λ/n\bar\lambda=\lambda/nλˉ=λ/n. Hellinger squared is ∑i(ai−bi)2\sum_i(\sqrt{a_i}-\sqrt{b_i})^2∑i​(ai​​−bi​​)2, chi-squared is ∑i(ai−bi)2/bi\sum_i(a_i-b_i)^2/b_i∑i​(ai​−bi​)2/bi​, and KL is ∑iailn⁡(ai/bi)\sum_i a_i\ln(a_i/b_i)∑i​ai​ln(ai​/bi​). The estimate is the first argument, zero-over-zero terms contribute zero, and positive estimated mass at zero true mass gives infinite divergence. A superscript (k)(k)(k) retains only the first kkk terms, without renormalization.

dH2(a,b)≤dχ2(a,b),dH2(k)(a,b)≤dχ2(k)(a,b),dKL(a,b)≤dχ2(a,b).d_{H^2}(a,b)\le d_{\chi^2}(a,b),\qquad d_{H^2}^{(k)}(a,b)\le d_{\chi^2}^{(k)}(a,b),\qquad d_{\mathrm{KL}}(a,b)\le d_{\chi^2}(a,b).dH2​(a,b)≤dχ2​(a,b),dH2(k)​(a,b)≤dχ2(k)​(a,b),dKL​(a,b)≤dχ2​(a,b).

This isolates the source comparison or construction with all of its domain and occurrence conventions retained.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Extend50_P32
Formal statement
noncomputable section

open MeasureTheory Filter

open scoped BigOperators ENNReal Matrix ComplexOrder Topology

namespace QAlgorithms.Papers.Independent.Extend50.P32

open QAlgorithms.Papers.Independent.Support50.P38 (wordMass tv)

-- For the density matrices below, the characteristic-polynomial identity
-- preserves all eigenvalues with multiplicity and is equivalent to unitary
-- diagonalization with these real eigenvalues. No accuracy is assumed.

open QAlgorithms.Papers.Independent.Gap20.P17 (HasSpectrum)

/-- Proposition 2.7, PDF p.12: full/full, truncated/truncated, and full KL/chi.
No probability support is excluded. -/
theorem classicalLossComparisons {d : ℕ} (hd : 0 < d) (alpha beta : Fin d → ℝ)
    (hAlpha : Learning.IsDistribution alpha) (hBeta : Learning.IsDistribution beta)
    (k : ℕ) (hk : 1 ≤ k) (hkd : k ≤ d) :
    ENNReal.ofReal (h2 d alpha beta) ≤ chi2 d alpha beta ∧
    ENNReal.ofReal (h2 k alpha beta) ≤ chi2 k alpha beta ∧
    kl alpha beta ≤ chi2 d alpha beta := by sorry

end QAlgorithms.Papers.Independent.Extend50.P32

end
Source
Efficient quantum tomography II, arXiv:1612.00034v1, PDF p. 12, Proposition 2.7
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

Let d>0d>0d>0, let α,β∈Rd\alpha,\beta\in\mathbb R^dα,β∈Rd have nonnegative coordinates summing to one, and let 1≤k≤d1\le k\le d1≤k≤d. For 0≤r≤d0\le r\le d0≤r≤d, define

Hr(α,β)=∑i<r(αi−βi)2,Xr(α,β)=∑i<r{0,βi=0, αi=0,+∞,βi=0, αi≠0,max⁡{(αi−βi)2/βi,0},βi≠0.H_r(\alpha,\beta)=\sum_{i<r}(\sqrt{\alpha_i}-\sqrt{\beta_i})^2, \qquad X_r(\alpha,\beta)=\sum_{i<r} \begin{cases} 0,&\beta_i=0,\ \alpha_i=0,\\ +\infty,&\beta_i=0,\ \alpha_i\ne0,\\ \max\{(\alpha_i-\beta_i)^2/\beta_i,0\},&\beta_i\ne0. \end{cases}Hr​(α,β)=i<r∑​(αi​​−βi​​)2,Xr​(α,β)=i<r∑​⎩⎨⎧​0,+∞,max{(αi​−βi​)2/βi​,0},​βi​=0, αi​=0,βi​=0, αi​=0,βi​=0.​

Define K(α,β)=+∞K(\alpha,\beta)=+\inftyK(α,β)=+∞ if some coordinate satisfies αi≠0\alpha_i\ne0αi​=0 and βi=0\beta_i=0βi​=0; otherwise define

K(α,β)=max⁡ ⁣{∑i<d{0,αi=0,αilog⁡(αi/βi),αi≠0, 0},K(\alpha,\beta)= \max\!\left\{ \sum_{i<d} \begin{cases} 0,&\alpha_i=0,\\ \alpha_i\log(\alpha_i/\beta_i),&\alpha_i\ne0, \end{cases} \,0\right\},K(α,β)=max{i<d∑​{0,αi​log(αi​/βi​),​αi​=0,αi​=0,​0},

using the natural logarithm and viewing finite nonnegative values as extended nonnegative reals. The three assertions are

max⁡{Hd(α,β),0}≤Xd(α,β),max⁡{Hk(α,β),0}≤Xk(α,β),K(α,β)≤Xd(α,β).\max\{H_d(\alpha,\beta),0\}\le X_d(\alpha,\beta), \qquad \max\{H_k(\alpha,\beta),0\}\le X_k(\alpha,\beta), \qquad K(\alpha,\beta)\le X_d(\alpha,\beta).max{Hd​(α,β),0}≤Xd​(α,β),max{Hk​(α,β),0}≤Xk​(α,β),K(α,β)≤Xd​(α,β).

Zero coordinates are permitted, with the infinite-value cases specified above; no sorting assumption is made.

Human review
  • Endorsed by Shuze Chen · Oct 29, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 29, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me