Documentation

Crn.JumpChain

CRN-1 as a statement about jump chains #

The dictionary relation jump-chain ("the same chain once idle steps are removed") is Dynamics.Kernel.SameJumpChain: the two kernels agree once each is conditioned on moving (Dynamics.Kernel.jump). On count vectors, one interaction of the population protocol of a network N is a kernel (ppStepKernel) that stays put when the drawn pair does not react. Its jump chain is the jump chain of stochastic mass-action kinetics with a common rate constant (ppStepKernel_jump), which is its own jump chain (Network.jumpKernel_jump): the protocol and the network have the same jump chain. Conditioning on the drawn pair reacting, as in ppKernel, is conditioning on moving (ppStepKernel_jump_eq_ppKernel), so this is CRN-1 (jumpKernel_eq_ppKernel) in the vocabulary of Dynamics.Kernel; for the approximate-majority network it is ApproxMajority.network_sameJumpChain.

noncomputable def Crn.ppStepKernel {S : Type u_1} [Fintype S] [DecidableEq S] {n : ℕ} (N : Network S) (hn : 2 ≤ n) :

One interaction of the population protocol of N, as a kernel on count vectors: from x, one interaction from some configuration with counts x (by ppStep_expect, the result depends on the counts only).

Equations
Instances For
    theorem Crn.Network.jumpKernel_jump {S : Type u_1} [Fintype S] [DecidableEq S] {n : ℕ} (N : Network S) {k : ℝ} (hk : 0 < k) :
    (N.jumpKernel k hk n).jump = N.jumpKernel k hk n

    The jump chain of stochastic mass-action kinetics is its own jump chain (in the sense of Dynamics.Kernel.jump): it stays put only at terminal states.

    theorem Crn.ppStepKernel_jump {S : Type u_1} [Fintype S] [DecidableEq S] {n : ℕ} (N : Network S) {k : ℝ} (hk : 0 < k) (hn : 2 ≤ n) :
    (ppStepKernel N hn).jump = N.jumpKernel k hk n

    CRN-1, jump-chain form. The jump chain of the population protocol of N on count vectors (an interaction conditioned on changing the counts) is the jump chain of stochastic mass-action kinetics with a common rate constant k.

    theorem Crn.ppStepKernel_jump_eq_ppKernel {S : Type u_1} [Fintype S] [DecidableEq S] {n : ℕ} (N : Network S) (hn : 2 ≤ n) :

    The protocol conditioned on the drawn pair reacting (ppKernel) is the jump chain of the protocol.

    theorem Crn.ppStepKernel_sameJumpChain {S : Type u_1} [Fintype S] [DecidableEq S] {n : ℕ} (N : Network S) {k : ℝ} (hk : 0 < k) (hn : 2 ≤ n) :

    The population protocol of N and stochastic mass-action kinetics have the same jump chain (the relation jump-chain of the dictionary, row population-protocols).

    theorem Crn.ApproxMajority.network_sameJumpChain {k : ℝ} (hk : 0 < k) {n : ℕ} (hn : 2 ≤ n) :

    CRN-1 for approximate majority, jump-chain form (network_jumpKernel_eq_ppKernel with Dynamics.Kernel.SameJumpChain). The population protocol of the approximate-majority network on count vectors and the approximate-majority CRN with rate constant k have the same jump chain: once the interactions in which the drawn pair does not react are removed, the protocol is the CRN's jump chain.