Peak estimation

Tim Taylor

library(outbreaks)
library(incidence2)
library(i2extras)

Bootstrapping and finding peaks

We provide functions to return the peak of the incidence data (grouped or ungrouped), bootstrap from the incidence data, and estimate confidence intervals around a peak.

bootstrap()

dat <- fluH7N9_china_2013
x <- incidence(dat, date_index = "date_of_onset", groups = "gender")
bootstrap(x)
#> # incidence:  67 x 4
#> # count vars: date_of_onset
#> # groups:     gender
#>    date_index gender count_variable count
#>  * <date>     <fct>  <chr>          <int>
#>  1 2013-02-19 m      date_of_onset      3
#>  2 2013-02-27 m      date_of_onset      1
#>  3 2013-03-07 m      date_of_onset      2
#>  4 2013-03-08 m      date_of_onset      1
#>  5 2013-03-09 f      date_of_onset      3
#>  6 2013-03-13 f      date_of_onset      1
#>  7 2013-03-17 m      date_of_onset      2
#>  8 2013-03-19 f      date_of_onset      3
#>  9 2013-03-20 f      date_of_onset      2
#> 10 2013-03-20 m      date_of_onset      4
#> # … with 57 more rows

find_peak()

dat <- fluH7N9_china_2013
x <- incidence(dat, date_index = "date_of_onset", groups = "gender")

# peaks across each group
find_peak(x)
#> # incidence:  2 x 4
#> # count vars: date_of_onset
#> # groups:     gender
#>   date_index gender count_variable count
#> * <date>     <fct>  <chr>          <int>
#> 1 2013-04-11 f      date_of_onset      3
#> 2 2013-04-03 m      date_of_onset      6

# peak without groupings
find_peak(regroup(x))
#> # incidence:  1 x 3
#> # count vars: date_of_onset
#>   date_index count_variable count
#> * <date>     <chr>          <int>
#> 1 2013-04-03 date_of_onset      7

estimate_peak()

Note that the bootstrapping approach used for estimating the peak time makes the following assumptions:

dat <- fluH7N9_china_2013
x <- incidence(dat, date_index = "date_of_onset", groups = "province")

# regrouping for overall peak (we suspend progress bar for markdown)
estimate_peak(regroup(x), progress = FALSE)
#> # A data frame: 1 × 7
#>   count_variable observed_peak observ…¹ boots…² lower_ci   median     upper_ci  
#>   <chr>          <date>           <int> <list>  <date>     <date>     <date>    
#> 1 date_of_onset  2013-04-03           7 <df>    2013-03-29 2013-04-08 2013-04-17
#> # … with abbreviated variable names ¹​observed_count, ²​bootstrap_peaks

# across provinces
estimate_peak(x, progress = FALSE)
#> # A data frame: 13 × 8
#>    province  count…¹ observed…² obser…³ boots…⁴ lower_ci   median     upper_ci  
#>    <fct>     <chr>   <date>       <int> <list>  <date>     <date>     <date>    
#>  1 Anhui     date_o… 2013-03-09       1 <df>    2013-03-09 2013-03-28 2013-04-14
#>  2 Beijing   date_o… 2013-04-11       1 <df>    2013-02-19 2013-04-11 2013-05-21
#>  3 Fujian    date_o… 2013-04-17       1 <df>    2013-04-17 2013-04-18 2013-04-29
#>  4 Guangdong date_o… 2013-07-27       1 <df>    2013-02-19 2013-07-27 2013-07-27
#>  5 Hebei     date_o… 2013-07-10       1 <df>    2013-02-19 2013-07-10 2013-07-10
#>  6 Henan     date_o… 2013-04-06       1 <df>    2013-04-06 2013-04-08 2013-04-17
#>  7 Hunan     date_o… 2013-04-14       1 <df>    2013-02-19 2013-04-14 2013-04-23
#>  8 Jiangsu   date_o… 2013-03-19       2 <df>    2013-03-08 2013-03-21 2013-04-19
#>  9 Jiangxi   date_o… 2013-04-15       1 <df>    2013-04-15 2013-04-17 2013-05-03
#> 10 Shandong  date_o… 2013-04-16       1 <df>    2013-02-19 2013-04-16 2013-04-27
#> 11 Shanghai  date_o… 2013-04-01       4 <df>    2013-03-17 2013-04-01 2013-04-10
#> 12 Taiwan    date_o… 2013-04-12       1 <df>    2013-02-19 2013-04-12 2013-04-12
#> 13 Zhejiang  date_o… 2013-04-06       5 <df>    2013-03-29 2013-04-08 2013-04-15
#> # … with abbreviated variable names ¹​count_variable, ²​observed_peak,
#> #   ³​observed_count, ⁴​bootstrap_peaks