Skip to main content
. Author manuscript; available in PMC: 2023 Jun 2.
Published in final edited form as: Ann Appl Stat. 2023 Jan 24;17(1):1–22. doi: 10.1214/21-aoas1583

Algorithm 3.

Updating rule in the ODE-based MCMC algorithm

1: Input: Parameter values from the previous interation I0,R0,γ,δ1:T,σ, geneology g. Proposal density q1(), q2() for updating the initial number of infected individuals and the removal rate.
2: Output Updated parameters values
3: Calculate X0:T,θ0:T based on I0,R0,γ,δ1:T,σ.
4: Propose I0 based on q1I0, then X0:T will be deterministically updated to X0:T according to I0,R0,γ,δ1:T,σ based on ODE integration.
5: Accept I0,XD:T with acceptance probability
amin1,Prgθ0:T,X0:TPrI0q1I0I0Prgθ0:T,XD:TPrI0q1I0I0
6: Propose γ based on q2γ, then X0:T,θ0:T will be deterministically updated to XD:T,θ0:T according to I0,R0,γ,δ1:T,σ
7: Accept γ,X0:T,θ0:T with acceptance probability
amin1,Prgθ0:T,X0:TPrγq2γγPrgθ0:T,X0:TPr(γ)q2γγ.
8: Let U=logR0,δ1:T,log(σ), then U a priori follows a multivariate normal distribution. Use elliptical slice sampler of obtain U and get the updated R0,δ1:T and σ. XD:T will be deterministically updated to X0:T according to I0,R0,γ,δ1:T,σ based on ODE integration.