Create Initial Conditions Generating Function — create_initial

Uses the output of create_stan_data() to create a function which can be used to sample from the prior distributions (or as close as possible) for parameters. Used in order to initialise each stan chain within a range of plausible values.

Usage

create_initial_conditions(stan_data, params)

Arguments

stan_data: A list of data as produced by create_stan_data().
params: A list of <EpiNow2.params> as created by make_param()

Value

An initial condition generating function

Details

R_mean is seeded from the initial-Rt prior carried in stan_data by make_init_priors(), so chains start near the configured reproduction number; this is a stopgap until derived-prior parameters are initialised through the shared path (#1481). The distribution code follows pack_init_prior() (0: lognormal, 1: gamma, 2: normal).

Infection model fields (fixed, estimate_r, bp_n and week_effect) may be absent, as in the secondary and truncation models. Their components are then initialised as empty.