Bootstrap incidence time series

This function can be used to bootstrap incidence objects. Bootstrapping is done by sampling with replacement the original input dates. See details for more information on how this is implemented.

bootstrap(x, randomise_groups = FALSE)

Arguments

x	An incidence object.
randomise_groups	A logical indicating whether groups should be randomised as well in the resampling procedure; respective group sizes will be preserved, but this can be used to remove any group-specific temporal dynamics. If FALSE (default), data are resampled within groups.

Value

An incidence object.

Details

As original data are not stored in incidence objects, the bootstrapping is achieved by multinomial sampling of date bins weighted by their relative incidence.

See also

find_peak to use estimate peak date using bootstrap

Examples

if (require(outbreaks) && require(ggplot2)) { withAutoprint({
  i <- incidence(fluH7N9_china_2013$date_of_onset)
  i
  plot(i)

  ## one simple bootstrap
  x <- bootstrap(i)
  x
  plot(x)

})}
#> Loading required package: outbreaks
#> Loading required package: ggplot2
#> > i <- incidence(fluH7N9_china_2013$date_of_onset)
#> 10 missing observations were removed.
#> > i
#> <incidence object>
#> [126 cases from days 2013-02-19 to 2013-07-27]
#> 
#> $counts: matrix with 159 rows and 1 columns
#> $n: 126 cases in total
#> $dates: 159 dates marking the left-side of bins
#> $interval: 1 day
#> $timespan: 159 days
#> $cumulative: FALSE
#> 
#> > plot(i)
#> > x <- bootstrap(i)
#> > x
#> <incidence object>
#> [126 cases from days 2013-02-19 to 2013-05-21]
#> 
#> $counts: matrix with 92 rows and 1 columns
#> $n: 126 cases in total
#> $dates: 92 dates marking the left-side of bins
#> $interval: 1 day
#> $timespan: 92 days
#> $cumulative: FALSE
#> 
#> > plot(x)