Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 3.4 — truncating a reversible process

Proved
KellyStochasticNetworks.truncated_reversible

by naimengye · Sep 18, 2026 · Mathlib 0df444a (Lean v4.33.1)

loss-networksmarkov-chainreversibilitystochastic-networks

Let qqq be the transition rates of a Markov process on a state space SSS, and let π=(π(j))j∈S\pi = (\pi(j))_{j \in S}π=(π(j))j∈S​ be in detailed balance with qqq, so that the process is reversible with equilibrium distribution π\piπ. Truncate the process to a subset A⊆SA \subseteq SA⊆S: keep the rates between states of AAA unchanged and suppress every transition that would leave AAA.

Then the truncated process is again reversible, with equilibrium distribution

π(j)(∑k∈Aπ(k))−1,j∈A.\pi(j)\Bigl(\sum_{k \in A}\pi(k)\Bigr)^{-1}, \qquad j \in A .π(j)(k∈A∑​π(k))−1,j∈A.

Precisely, if the sum Z=∑k∈Aπ(k)Z = \sum_{k \in A}\pi(k)Z=∑k∈A​π(k) converges and is non-zero, then j↦π(j)/Zj \mapsto \pi(j)/Zj↦π(j)/Z is in detailed balance with the truncated rates and its terms sum to 111.

This is the lever of the whole chapter. A loss network is the truncation of a system of independent, uncapacitated queues to the feasible set {n:An≤C}\{n : An \le C\}{n:An≤C}, and because the uncapacitated system is reversible the lemma hands over its equilibrium distribution at once: independent Poisson random variables conditioned on a set of linear inequalities.

Formalization Note The truncated process lives on the subtype of states belonging to AAA, so the suppression of transitions leaving AAA is automatic: such a transition has no target. Convergence of ZZZ and its value are asserted together.

Preamble
import Mathlib
import Definitions.Def_KellyStochasticNetworks_Balance
import Definitions.Def_KellyStochasticNetworks_Erlang
import Definitions.Def_KellyStochasticNetworks_Migration
import Definitions.Def_KellyStochasticNetworks_LossNetwork
Formal statement
namespace KellyStochasticNetworks

theorem truncated_reversible {S : Type*} (π : S → ℝ) (q : S → S → ℝ) (A : Set S)
    (h : DetailedBalance π q) (Z : ℝ) (hZ0 : Z ≠ 0)
    (hZ : HasSum (fun j : A => π (j : S)) Z) :
    DetailedBalance (fun j : A => π (j : S) / Z) (truncatedRates q A)
      ∧ HasSum (fun j : A => π (j : S) / Z) 1 := by sorry

end KellyStochasticNetworks
Source
Kelly & Yudovina, Stochastic Networks, CUP 2014, p. 53 (PDF p. 61), Lemma 3.4: 'If a reversible Markov process with state space S and equilibrium distribution (pi(j), j in S) is truncated to A subset of S, the resulting Markov process is reversible and has equilibrium distribution pi(j) (sum_{k in A} pi(k))^{-1}, j in A. (3.2)' sha256 ec271d555059aee58613e5e9a98b8346214b16185c527314f5d94c1ac8b17b6a
Human review
  • Endorsed by Shuze Chen · Sep 25, 2026

    Confirmed by the moderator at approval.

  • Endorsed by naimengye · Sep 25, 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