Convert simulated event times to dates

Takes the continuous event times produced by simulate_gillespie() and simulate_secondary() and returns the dates an analyst would actually receive. Event times are floored to their reporting window, so each event is known only by the window it fell in, and are then offset from outbreak_start_date.

Usage

simulate_dates(
  data,
  outbreak_start_date = as.Date("2024-01-01"),
  primary_window = 1,
  secondary_window = primary_window,
  obs_time = NULL,
  keep_times = FALSE
)

Arguments

data

A data.frame with numeric ptime and stime columns, as returned by simulate_secondary().

outbreak_start_date

The date the outbreak started, corresponding to time zero.

primary_window

Width of the primary event reporting window in days. Either a single value used for every observation, or one value per row of data. The default of 1 gives daily reporting. Use 7 for weekly.

secondary_window

Width of the secondary event reporting window in days, in the same form as primary_window. Defaults to primary_window, so the two events share a reporting interval unless you say otherwise.

obs_time

Optional numeric observation time, in the same units as ptime and stime. When supplied an obs_date column is added. When NULL, the default, no observation date is added and as_epidist_linelist_data() falls back to the day after the last secondary event.

keep_times

Whether to keep the underlying numeric times. Useful when comparing estimates against the values used to simulate.

Value

A data.frame with pdate_lwr, pdate_upr, sdate_lwr and sdate_upr columns, and obs_date when obs_time is supplied.

Details

The returned columns are named to match as_epidist_linelist_data(), so the output can be passed straight to it.

See also

Other simulate: simulate_exponential_cases(), simulate_gillespie(), simulate_secondary(), simulate_uniform_cases()

Examples

simulate_gillespie(seed = 1) |>
  simulate_secondary(meanlog = 1.8, sdlog = 0.5) |>
  simulate_dates(outbreak_start_date = as.Date("2024-02-01")) |>
  head()
#>   case  pdate_lwr  pdate_upr  sdate_lwr  sdate_upr
#> 1    1 2024-02-01 2024-02-02 2024-02-09 2024-02-10
#> 2    2 2024-02-01 2024-02-02 2024-02-06 2024-02-07
#> 3    3 2024-02-01 2024-02-02 2024-02-09 2024-02-10
#> 4    4 2024-02-01 2024-02-02 2024-02-10 2024-02-11
#> 5    5 2024-02-01 2024-02-02 2024-02-08 2024-02-09
#> 6    6 2024-02-01 2024-02-02 2024-02-10 2024-02-11