. Author manuscript; available in PMC: 2021 Jul 30.

Published in final edited form as: Stat Methods Med Res. 2020 Jan 30;29(9):2520–2537. doi: 10.1177/0962280219889080

Table A1:

Parameter constraints for the L-BFGS-B algorithm.

Scenario	Candidate Distribution	θ ₁	θ ₂
S ₁	Normal	μ ∈ (Q_min, Q_max)	σ ∈ (10⁻³, 50)
	Log-Normal	μ ∈ (log(Q_min), log(Q_max))	σ ∈ (10⁻³, 50)
	Gamma	α ∈ (10⁻³, 100)	β ∈ (10⁻³, 100)
	Beta	α ∈ (10⁻³, 40)	β ∈ (10⁻³, 40)
	Weibull	λ ∈ (10⁻³, 100)	k ∈ (10⁻³, 100)
S₂ & S₃	Normal	μ ∈ (Q₁, Q₃)	σ ∈ (10⁻³, 50)
	Log-Normal	μ ∈ (log(Q₁), log(Q₃))	σ ∈ (10⁻³, 50)
	Gamma	α ∈ (10⁻³, 100)	β ∈ (10⁻³, 100)
	Beta	α ∈ (10⁻³, 40)	β ∈ (10⁻³, 40)
	Weibull	λ ∈ (10⁻³, 100)	k ∈ (10⁻³, 100)