Date Conversions

MMWR Conversions

For most time series data, we prefer to aggregate at the week or year level for reporting purposes. Like most epidemiologists, we rely on MMWR disease week and year. For week, we have two flavors - mwwr_week and week_ending_date (Saturday of disease week). For year, we use mmwr_year.

dates <- linelist |>
  select(SpecimenDate) |>
  arrange(SpecimenDate) |>
  mutate(
    epi_year = mmwr_year(SpecimenDate),
    disease_week = mmwr_week(SpecimenDate),
    week_ending = week_ending_date(SpecimenDate)
  )

head(dates)
#> # A tibble: 6 × 4
#>   SpecimenDate epi_year disease_week week_ending
#>   <date>          <dbl>        <dbl> <date>     
#> 1 2022-06-04       2022           22 2022-06-04 
#> 2 2022-06-04       2022           22 2022-06-04 
#> 3 2022-06-04       2022           22 2022-06-04 
#> 4 2022-06-05       2022           23 2022-06-11 
#> 5 2022-06-05       2022           23 2022-06-11 
#> 6 2022-06-06       2022           23 2022-06-11

mmwrweek_to_date()

For those that need to convert epidemiological year and week to week ending date:

dates <- dates |>
  select(epi_year, disease_week) |>
  mutate(
    week_ending = mmwrweek_to_date(epi_year, disease_week)
  )

head(dates)
#> # A tibble: 6 × 3
#>   epi_year disease_week week_ending
#>      <dbl>        <dbl> <date>     
#> 1     2022           22 2022-06-04 
#> 2     2022           22 2022-06-04 
#> 3     2022           22 2022-06-04 
#> 4     2022           23 2022-06-11 
#> 5     2022           23 2022-06-11 
#> 6     2022           23 2022-06-11

mmwr_calendar()

To make mmwr_calendar() with total disease weeks + start/end dates:

mmwr_calendar(2023) |>
  head(20)
#>    Year Week      Start        End
#> 1  2023    1 2023-01-01 2023-01-07
#> 2  2023    2 2023-01-08 2023-01-14
#> 3  2023    3 2023-01-15 2023-01-21
#> 4  2023    4 2023-01-22 2023-01-28
#> 5  2023    5 2023-01-29 2023-02-04
#> 6  2023    6 2023-02-05 2023-02-11
#> 7  2023    7 2023-02-12 2023-02-18
#> 8  2023    8 2023-02-19 2023-02-25
#> 9  2023    9 2023-02-26 2023-03-04
#> 10 2023   10 2023-03-05 2023-03-11
#> 11 2023   11 2023-03-12 2023-03-18
#> 12 2023   12 2023-03-19 2023-03-25
#> 13 2023   13 2023-03-26 2023-04-01
#> 14 2023   14 2023-04-02 2023-04-08
#> 15 2023   15 2023-04-09 2023-04-15
#> 16 2023   16 2023-04-16 2023-04-22
#> 17 2023   17 2023-04-23 2023-04-29
#> 18 2023   18 2023-04-30 2023-05-06
#> 19 2023   19 2023-05-07 2023-05-13
#> 20 2023   20 2023-05-14 2023-05-20

assign_season()

For epidemiologists working viral respiratory surveillance, patients/laboratory results can also be categorized by season. Season in this context spans week 40 of current year to week 39 of following year.

df <- data.frame(spec_date = as.Date(c("2023-10-01","2023-11-04","2024-09-28","2024-09-29")))

df |>
  mutate(season = assign_season(spec_date))
#>    spec_date  season
#> 1 2023-10-01 2023-24
#> 2 2023-11-04 2023-24
#> 3 2024-09-28 2023-24
#> 4 2024-09-29 2024-25

to_month()

An alternative to grouping cases at the year or week level is by month. Returned output is a date formatted YYYY-MM-01.

df <- data.frame(spec_date = as.Date(c("2023-10-01","2023-11-04","2024-09-28","2024-09-29")))

df |>
  mutate(month = to_month(spec_date))
#>    spec_date      month
#> 1 2023-10-01 2023-10-01
#> 2 2023-11-04 2023-11-01
#> 3 2024-09-28 2024-09-01
#> 4 2024-09-29 2024-09-01