Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 2 — Algorithm 2 in N∞(β)\mathcal N_\infty(\beta)N∞​(β) or N∞−(β)\mathcal N^-_\infty(\beta)N∞−​(β) with γ≤2(1−β)\gamma\le2(1-\beta)γ≤2(1−β) reaches precision ttt in ⌈ntln⁡2/(2βγ(1−γ))⌉\lceil nt\ln 2/(2\beta\gamma(1-\gamma))\rceil⌈ntln2/(2βγ(1−γ))⌉ steps

Proved
AdaptiveStepIPM.WideNbhd.theorem_2

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

complexityinterior-point-methodslinear-programmingp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Consider the linear program (P) min⁡cTx\min c^TxmincTx s.t. Ax=bAx=bAx=b, x≥0x\ge0x≥0, and its dual (D), with n≥1n\ge1n≥1 variables. Let β∈(0,1)\beta\in(0,1)β∈(0,1) and γ∈(0,1)\gamma\in(0,1)γ∈(0,1) be constants with γ≤2(1−β)\gamma\le 2(1-\beta)γ≤2(1−β), and let N\mathcal NN be N∞(β)\mathcal N_\infty(\beta)N∞​(β) or N∞−(β)\mathcal N^-_\infty(\beta)N∞−​(β).

Algorithm 2. Given (x0,s0)∈N(x^0,s^0)\in\mathcal N(x0,s0)∈N with (x0)Ts0≤2t(x^0)^Ts^0\le 2^t(x0)Ts0≤2t, set k=0k = 0k=0. While (xk)Tsk>2−t(x^k)^Ts^k > 2^{-t}(xk)Tsk>2−t: compute d=d(xk,sk,γ)d = d(x^k,s^k,\gamma)d=d(xk,sk,γ) from (2), compute the largest θˉ\bar\thetaθˉ such that (x(θ),s(θ))∈N(x(\theta),s(\theta))\in\mathcal N(x(θ),s(θ))∈N for all θ∈[0,θˉ]\theta\in[0,\bar\theta]θ∈[0,θˉ], set (xk+1,sk+1)=(x(θˉ),s(θˉ))(x^{k+1},s^{k+1}) = (x(\bar\theta),s(\bar\theta))(xk+1,sk+1)=(x(θˉ),s(θˉ)) and k=k+1k = k+1k=k+1.

Theorem. Every run of Algorithm 2 reaches (xk)Tsk≤2−t(x^k)^Ts^k\le2^{-t}(xk)Tsk≤2−t for some

k  ≤  K:=⌈ln⁡2⋅n t2βγ(1−γ)⌉,k \;\le\; K := \Big\lceil \frac{\ln 2\cdot n\,t}{2\beta\gamma(1-\gamma)}\Big\rceil ,k≤K:=⌈2βγ(1−γ)ln2⋅nt​⌉,

so the algorithm terminates in O(nt)O(nt)O(nt) iterations.

This is the classical O(nt)O(nt)O(nt) bound of Kojima, Mizuno and Yoshise, obtained for an adaptive-step method in the wide neighbourhoods, which with β\betaβ close to 111 cover almost all strictly feasible pairs.

Formalization Note The paper states "O(nt)O(nt)O(nt) iterations"; KKK is the explicit count that its inequality (15), μk+1≤(1−4βγ(1−γ)/n)μk\mu^{k+1}\le(1-4\beta\gamma(1-\gamma)/n)\mu^kμk+1≤(1−4βγ(1−γ)/n)μk, gives: (xk)Tsk≤e−4kβγ(1−γ)/n 2t≤2−t(x^k)^Ts^k\le e^{-4k\beta\gamma(1-\gamma)/n}\,2^t\le 2^{-t}(xk)Tsk≤e−4kβγ(1−γ)/n2t≤2−t once k≥Kk\ge Kk≥K. A run is any pair of sequences (xk,sk)(x^k,s^k)(xk,sk) that starts as required and makes one Algorithm 2 iteration from every iterate with (xk)Tsk>2−t(x^k)^Ts^k>2^{-t}(xk)Tsk>2−t; what happens after the stopping test is met is unconstrained. "The largest θˉ\bar\thetaθˉ" is encoded as a greatest element, so a run exists only while such a maximum exists (for n=1n=1n=1 it never does, and the theorem holds trivially). No sign condition on ttt is imposed: for t≤0t\le0t≤0 the starting pair already satisfies (x0)Ts0≤2t≤2−t(x^0)^Ts^0\le2^t\le2^{-t}(x0)Ts0≤2t≤2−t and K=0K=0K=0. The two neighbourhoods are handled by a hypothesis N=N∞(β)\mathcal N = \mathcal N_\infty(\beta)N=N∞​(β) or N=N∞−(β)\mathcal N = \mathcal N^-_\infty(\beta)N=N∞−​(β).

Preamble
import Mathlib
import Definitions.Def_AdaptiveStepIPM_WideNbhd_Neighborhoods
import Definitions.Def_AdaptiveStepIPM_WideNbhd_Algorithm2
Formal statement
namespace AdaptiveStepIPM.WideNbhd

open Matrix

/-- Theorem 2 of Mizuno–Todd–Ye (p. 10): let `β, γ ∈ (0, 1)` with `γ ≤ 2(1 − β)`. Algorithm 2 with
`N = N_∞(β)` or `N_∞⁻(β)`, started at `(x⁰, s⁰) ∈ N` with `(x⁰)ᵀs⁰ ≤ 2^t`, reaches
`(x^k)ᵀs^k ≤ 2^{−t}` within `K = ⌈ln 2 · n · t / (2βγ(1 − γ))⌉` iterations — the explicit count
behind the paper's `O(nt)`, from (15). A run is any sequence of pairs that starts as required
and makes an Algorithm 2 step from every iterate with `(x^k)ᵀs^k > 2^{−t}`. -/
theorem theorem_2 {n m : ℕ} (hn : 1 ≤ n) (A : Matrix (Fin m) (Fin n) ℝ) (b : Fin m → ℝ)
    (c : Fin n → ℝ) (β γ : ℝ) (hβ0 : 0 < β) (hβ1 : β < 1) (hγ0 : 0 < γ) (hγ1 : γ < 1)
    (hγβ : γ ≤ 2 * (1 - β)) (N : Set ((Fin n → ℝ) × (Fin n → ℝ)))
    (hN : N = Ninf A b c β ∨ N = NinfMinus A b c β) (t : ℝ) (x s : ℕ → Fin n → ℝ)
    (h0 : (x 0, s 0) ∈ N) (h0t : x 0 ⬝ᵥ s 0 ≤ (2 : ℝ) ^ t)
    (hrun : ∀ k, (2 : ℝ) ^ (-t) < x k ⬝ᵥ s k → Alg2Step A N γ (x k) (s k) (x (k + 1)) (s (k + 1))) :
    ∃ k ≤ ⌈Real.log 2 * n * t / (2 * β * γ * (1 - γ))⌉₊, x k ⬝ᵥ s k ≤ (2 : ℝ) ^ (-t) := by sorry

end AdaptiveStepIPM.WideNbhd
Source
Mizuno, Todd, Ye, On adaptive-step primal-dual interior-point algorithms for linear programming, Cornell ORIE Technical Report No. 944 (1990, rev. 1991), p. 10, Theorem 2 (Algorithm 2, pp. 9–10)
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 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