concentration_zero
theoremverifiedimported Lean
Concentration bound at the zero vector. Let $\mu$ be a probability measure on a measurable space $\Omega$, let $\omega \mapsto A(\omega)$ be a family of $k \times d$ real matrices indexed by $\Omega$, and let $\varepsilon \in \bbr$. Then the probability that the zero vector $0 \in \bbr^d$ triggers the bad distortion event, namely that \[ \varepsilon\,\|0\|^2 < \bigl|\,\|A(\omega)\,0\|^2 - \|0\|^2\,\bigr|, \] satisfies \[ \mu\bigl\{\omega : \varepsilon\,\|0\|^2 < \bigl|\,\|A(\omega)\,0\|^2 - \|0\|^2\,\bigr|\bigr\} \;\le\; 2\exp\!\Bigl(-\tfrac{k\,\varepsilon^2}{8}\Bigr). \]
Statement
set_option relaxedAutoImplicit false in set_option autoImplicit false in open MeasureTheory ProbabilityTheory Real NNReal Matrix Finset in variable {d k : ℕ} in lemma concentration_zero {Ω : Type*} [MeasurableSpace Ω] (μ : Measure Ω) [IsProbabilityMeasure μ] (A : Ω → Matrix (Fin k) (Fin d) ℝ) (ε : ℝ) : (μ {ω | BadSingle ε (A ω) (0 : EuclideanSpace ℝ (Fin d))}).toReal ≤ 2 * Real.exp (-(k : ℝ) * ε ^ 2 / 8)
- Source
- TCSlib, `concentration_zero`
- Lean proof
- By Ganesh Sankar, from github.com/Shilun-Allan-Li/tcslib at
c8a02591b1e5(Copyright (c) 2026 Ganesh Sankar. All rights reserved, Apache-2.0). Compiled here against leanprover/lean4:v4.33.1, Mathlib v4.33.1; written for leanprover/lean4:v4.25.0.concentration_zero← concentration_zero (TCSlib/LearningTheory/JohnsonLindenstrauss/RowDistribution.lean:69): verbatim
- Verified
- 07 Oct 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 1
- BadSingleFor a parameter $\varepsilon\in\mathbb{R}$, a matrix $A$, and a vector $x\in\mathbb{R}^d$, the predicate…
- Used by 1
- jl_concentration_single_via_chi_squaredJohnson–Lindenstrauss concentration for a single vector. Let $k \ge 1$, and let $A$ be a random $k \times d$ real matrix on a probability…
Lean source view module on GitHub
Choose a tactic — a marked line — to see the goals Lean had before and after it, as in Lean’s infoview. Step with ← → or j k.
set_option relaxedAutoImplicit false inset_option autoImplicit false inopen MeasureTheory ProbabilityTheory Real NNReal Matrix Finset invariable {d k : ℕ} in/-- For the zero vector the single-vector concentration bound holds trivially: for any random matrix `A` and any `ε`, the probability of the bad event `BadSingle ε (A ω) 0` is at most `2·exp(−kε²/8)`, because that event is empty. -/lemma concentration_zero {Ω : Type*} [MeasurableSpace Ω] (μ : Measure Ω) [IsProbabilityMeasure μ] (A : Ω → Matrix (Fin k) (Fin d) ℝ) (ε : ℝ) : (μ {ω | BadSingle ε (A ω) (0 : EuclideanSpace ℝ (Fin d))}).toReal ≤ 2 * Real.exp (-(k : ℝ) * ε ^ 2 / 8) := by have hempty : {ω | BadSingle ε (A ω) (0 : EuclideanSpace ℝ (Fin d))} = ∅ := by ext ω simp only [BadSingle, Set.mem_setOf_eq, Set.mem_empty_iff_false, iff_false, not_lt] simp [map_zero] rw [hempty, measure_empty, ENNReal.toReal_zero] positivity