Relate Time Steps to Observed Data for mcstate — filter

This function formats observed data into a structure usable by the mcstate particle filter, adjusting for time steps (weekly or monthly) and ensuring compatibility with the filtering framework.

Usage

filter_data(incidence_observed, month = FALSE, initial_time_obs = 0, rate = 7)

Arguments

incidence_observed: A data frame of observed incidence data. Must include a column named week_no (if month = FALSE) or month_no (if month = TRUE), and a column named value for the observed incidence.
month: Logical. If TRUE, assumes data is monthly; otherwise, weekly.
initial_time_obs: Integer. The starting time index for monthly data (default is 0).
rate: Integer. The time step rate (e.g., 7 for weekly, 30 for monthly). Default is 7.

Value

A data frame formatted for use with mcstate::particle_filter, including time start and end steps.

Examples

if (requireNamespace("mcstate", quietly = TRUE)) {
  # Weekly example
  incidence_observed <- data.frame(
    week_no = 0:10,
    value = rpois(11, lambda = 5)
  )
  formatted_data <- filter_data(incidence_observed, month = FALSE, initial_time_obs = 0, rate = 7)
  head(formatted_data)

  # Monthly example
  incidence_observed_month <- data.frame(
    month_no = 0:10,
    value = rpois(11, lambda = 10)
  )
  formatted_month_data <- filter_data(incidence_observed_month, month = TRUE,
  initial_time_obs = 0, rate = 30)
  head(formatted_month_data)
}
#>   month_no_start month_no_end time_start time_end value
#> 2              1            2          0       30    11
#> 3              2            3         30       60    15
#> 4              3            4         60       90    10
#> 5              4            5         90      120     6
#> 6              5            6        120      150     8
#> 7              6            7        150      180    11