Section 3.3.2 -- stationarity of the Glauber dynamics
ProvedMarkovMixing.glauber_stationaryLet be a probability distribution on the space of configurations , which assign a spin from a finite set to each site of a finite nonempty set . The Glauber dynamics (single-site heat bath) for makes one step as follows: pick a site uniformly at random, and replace the spin at by a sample from conditioned on agreeing with the current configuration at every site other than .
The theorem asserts four things about the resulting transition matrix . First, all entries are nonnegative. Second, for every configuration in the support of (i.e. with ) the row of at sums to one — so restricted to the support, is a genuine Markov chain. (Positivity of is not assumed, and at a configuration of zero mass the conditioning that defines a step can be vacuous, so row sums are claimed only on the support.) Third, satisfies the detailed balance equations for all configurations — reversibility with respect to . Fourth, is a stationary distribution: for every . This is Exercise 3.2 (§3.3.2) of Levin–Peres–Wilmer.
import Definitions.Def_mm_mcmc
namespace MarkovMixing
/-- **§3.3.2, Exercise 3.2** (LPW): the Glauber dynamics for a distribution
`π` on configurations is reversible with respect to `π` and has stationary
distribution `π`; its entries are nonnegative and every row at a
configuration in the support of `π` sums to `1`. -/
theorem glauber_stationary {Vv S : Type*} [Fintype Vv] [DecidableEq Vv]
[Fintype S] [DecidableEq S] [Nonempty Vv]
(π : (Vv → S) → ℝ) (hπ : IsDist π) :
(∀ x y : Vv → S, 0 ≤ glauber π x y) ∧
(∀ x : Vv → S, 0 < π x → ∑ y, glauber π x y = 1) ∧
DetailedBalance (glauber π) π ∧
IsStationary (glauber π) π := by
sorry
end MarkovMixing