An introduction to incidence2

The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

1. What does it do?
2. Functions at a glance
3. Basic Usage
4. Support for tidy-select semantics
5. Working with incidence objects
6. Building on incidence2

1. What does it do?

incidence2 is an R package that implements functions to compute, handle and visualise incidence data. It aims to be intuitive to use for both interactive data exploration and as part of more robust outbreak analytic pipelines.

The package is based around objects of the namesake class, incidence2. These objects are a tibble subclass with some additional invariants. That is, an incidence2 object must:

have one column representing the date index (this does not need to be a Date object but must have an inherent ordering over time);
have one column representing the count variable (i.e. what is being counted) and one variable representing the associated count;
have zero or more columns representing groups;
not have duplicated rows with regards to the date, group and count variables.

2. Functions at a glance

To create and work with incidence2 objects we provide a number of functions:

incidence(): for the construction of incidence objects from both linelists and pre-aggregated data sets.
regroup(): regroup incidence from different groups into one global incidence time series.
incidence_() and regroup_(): These work similar to their aforementioned namesakes but also add support for tidy-select semantics in their arguments.
plot.incidence2(): generate simple plots with reasonable defaults.
cumulate(): calculate the cumulative incidence over time.
complete_dates(): ensure every possible combination of date and groupings is represented with a count.
keep_first(), keep_last(): keep the rows corresponding to the first (or last) set of grouped dates (ordered by time) from an incidence2 object.
keep_peaks(); keep the rows corresponding to the maximum count value for each grouping of an incidence2 object. A convenience wrapper around this, first_peak() keeps returns the earliest occurring peak row.
bootstrap_incidence(); sampling (with replacement and optional randomisation) from incidence2 objects.
estimate_peak(); estimate the peak of an epidemic curve using bootstrapped samples of the available data.
Accessors for underlying variables: get_date_index(), get_count_variable(), get_count_value(), get_groups(), get_count_value_name(), get_count_variable_name(), get_date_index_name() and get_group_names().
Methods for common base R generics:
- as.data.frame.incidence2()
- $<-.incidence2()
- [.incidence2()
- [<-.incidence2()
- names<-.incidence2()
- split.incidence2()
- rbind.incidence2()
Methods for generics from the wider R package ecosystem, including:
- mutate.incidence2()
- summarise.incidence2()
- nest.incidence2()
- as_tibble.incidence2()
- as.data.table.incidence2()

3. Basic Usage

Examples in the vignette utilise three different sets of data:

A synthetic linelist generated from a simulated Ebola Virus Disease (EVD) outbreak and available in the outbreaks package.
A pre-aggregated time-series of Covid cases, tests, hospitalisations, and deaths for UK regions that is included within the incidence2 package.
136 cases of influenza A H7N9 in China. Again available in the outbreaks package.

3.1 Computing incidence from a linelist

Broadly speaking, we refer to data with one row of observations (e.g. ‘Sex’, ‘Date of symptom onset’, ‘Date of Hospitalisation’) per individual as a linelist

library(incidence2)
#> Loading required package: grates

# linelist from the simulated ebola outbreak  (removing some missing entries)
ebola <- subset(outbreaks::ebola_sim_clean$linelist ,!is.na(hospital))
str(ebola)
#> 'data.frame':	4373 obs. of  11 variables:
#>  $ case_id                : chr  "d1fafd" "53371b" "f5c3d8" "0f58c4" ...
#>  $ generation             : int  0 1 1 2 3 3 2 3 4 3 ...
#>  $ date_of_infection      : Date, format: NA "2014-04-09" ...
#>  $ date_of_onset          : Date, format: "2014-04-07" "2014-04-15" ...
#>  $ date_of_hospitalisation: Date, format: "2014-04-17" "2014-04-20" ...
#>  $ date_of_outcome        : Date, format: "2014-04-19" NA ...
#>  $ outcome                : Factor w/ 2 levels "Death","Recover": NA NA 2 2 2 1 2 1 NA 2 ...
#>  $ gender                 : Factor w/ 2 levels "f","m": 1 2 1 1 1 1 2 2 1 1 ...
#>  $ hospital               : Factor w/ 5 levels "Connaught Hospital",..: 2 1 3 3 1 4 3 5 1 1 ...
#>  $ lon                    : num  -13.2 -13.2 -13.2 -13.2 -13.2 ...
#>  $ lat                    : num  8.47 8.46 8.48 8.45 8.46 ...

To compute daily incidence we pass to incidence() our linelist data frame as well as the name of a column in the data that we can use to index over time. Whilst we refer to this index as the date_index there is no restriction on it’s type, save the requirement that is has an inherent ordering.

(daily_incidence <- incidence(ebola, date_index = "date_of_onset"))
#> # incidence:  361 x 3
#> # count vars: date_of_onset
#>    date_index count_variable count
#>    <date>     <chr>          <int>
#>  1 2014-04-07 date_of_onset      1
#>  2 2014-04-15 date_of_onset      1
#>  3 2014-04-21 date_of_onset      2
#>  4 2014-04-26 date_of_onset      1
#>  5 2014-05-01 date_of_onset      2
#>  6 2014-05-03 date_of_onset      1
#>  7 2014-05-05 date_of_onset      1
#>  8 2014-05-06 date_of_onset      2
#>  9 2014-05-07 date_of_onset      2
#> 10 2014-05-08 date_of_onset      2
#> # ℹ 351 more rows

incidence2 also provides a simple plot method (see help("plot.incidence2")) built upon ggplot2.

plot(daily_incidence)

Bar chart of daily incidence covering the period April 2014 to April 2015 inclusive. The graph appears to peaks around September 2014.

plot of chunk unnamed-chunk-4

The daily data is quite noisy, so it may be worth grouping the dates prior to calculating the incidence. One way to do this is to utilise functions from the grates package. incidence2 depends on the grates package so all of it’s functionality is available directly to users. Here we use the as_isoweek() function to convert the ‘date of onset’ to an isoweek (a week starting on a Monday) before proceeding to calculate the incidence:

(weekly_incidence <- 
    ebola |>
    mutate(date_of_onset = as_isoweek(date_of_onset)) |> 
    incidence(date_index = "date_of_onset"))
#> # incidence:  56 x 3
#> # count vars: date_of_onset
#>    date_index count_variable count
#>    <isowk>    <chr>          <int>
#>  1 2014-W15   date_of_onset      1
#>  2 2014-W16   date_of_onset      1
#>  3 2014-W17   date_of_onset      3
#>  4 2014-W18   date_of_onset      3
#>  5 2014-W19   date_of_onset     11
#>  6 2014-W20   date_of_onset     13
#>  7 2014-W21   date_of_onset     11
#>  8 2014-W22   date_of_onset     15
#>  9 2014-W23   date_of_onset     13
#> 10 2014-W24   date_of_onset     13
#> # ℹ 46 more rows
plot(weekly_incidence, border_colour = "white")

Bar chart of weekly incidence covering 2014-W15 to 2015-W18 inclusive. The graph peaks at 2014-W38. The "descent" from the peak tapers off slower than the initial "ascent".

plot of chunk unnamed-chunk-5

As this sort of date grouping is often required we have chosen to integrate this within the incidence() function via the interval parameter. interval can take any of the following values:

‘day’ or ‘daily’ (mapping to Date objects);
‘week(s)’, ‘isoweek(s)’ or ‘weekly’ (mapping to grates_isoweek);
‘epiweek(s)’ (mapping to grates_epiweek);
‘month(s)’, ‘yearmonth(s)’ or ‘monthly’ (grates_yearmonth);
‘quarter(s)’, ‘yearquarter(s)’ or ‘quarterly’ (grates_yearquarter);
‘year(s)’ or ‘yearly’ (grates_year).

As an example, the following is equivalent to the weekly_incidence output above:

(dat <- incidence(ebola, date_index = "date_of_onset", interval = "isoweek"))
#> # incidence:  56 x 3
#> # count vars: date_of_onset
#>    date_index count_variable count
#>    <isowk>    <chr>          <int>
#>  1 2014-W15   date_of_onset      1
#>  2 2014-W16   date_of_onset      1
#>  3 2014-W17   date_of_onset      3
#>  4 2014-W18   date_of_onset      3
#>  5 2014-W19   date_of_onset     11
#>  6 2014-W20   date_of_onset     13
#>  7 2014-W21   date_of_onset     11
#>  8 2014-W22   date_of_onset     15
#>  9 2014-W23   date_of_onset     13
#> 10 2014-W24   date_of_onset     13
#> # ℹ 46 more rows
# check equivalent
identical(dat, weekly_incidence)
#> [1] TRUE

If we wish to aggregate by specified groups we can use the groups argument. For instance, to compute the weekly incidence by gender:

(weekly_incidence_gender <- incidence(
    ebola,
    date_index = "date_of_onset",
    groups = "gender",
    interval = "isoweek"
))
#> # incidence:  109 x 4
#> # count vars: date_of_onset
#> # groups:     gender
#>    date_index gender count_variable count
#>    <isowk>    <fct>  <chr>          <int>
#>  1 2014-W15   f      date_of_onset      1
#>  2 2014-W16   m      date_of_onset      1
#>  3 2014-W17   f      date_of_onset      2
#>  4 2014-W17   m      date_of_onset      1
#>  5 2014-W18   f      date_of_onset      3
#>  6 2014-W19   f      date_of_onset      8
#>  7 2014-W19   m      date_of_onset      3
#>  8 2014-W20   f      date_of_onset      4
#>  9 2014-W20   m      date_of_onset      9
#> 10 2014-W21   f      date_of_onset      6
#> # ℹ 99 more rows

For grouped data, the plot method will create a faceted plot across groups unless a fill variable is specified:

plot(weekly_incidence_gender, border_colour = "white", angle = 45)

Two bar charts (side by side) of weekly incidence covering 2014-W15 to 2015-W18 inclusive. Females are on the left, Males the right. The graphs peak between 2014-W35 and 2014-W45. The "descent" from the peak tapers off slower than the initial "ascent".

plot of chunk unnamed-chunk-8

plot(weekly_incidence_gender, border_colour = "white", angle = 45, fill = "gender")

plot of chunk unnamed-chunk-9

incidence() also supports multiple date inputs and allows renaming via the use of named vectors:

(weekly_multi_dates <- incidence(
    ebola,
    date_index = c(
        onset = "date_of_onset",
        infection = "date_of_infection"
    ), 
    interval = "isoweek",
    groups = "gender"
))
#> # incidence:  216 x 4
#> # count vars: onset, infection
#> # groups:     gender
#>    date_index gender count_variable count
#>    <isowk>    <fct>  <chr>          <int>
#>  1 2014-W15   f      onset              1
#>  2 2014-W15   m      infection          1
#>  3 2014-W16   f      infection          1
#>  4 2014-W16   m      onset              1
#>  5 2014-W17   f      infection          3
#>  6 2014-W17   f      onset              2
#>  7 2014-W17   m      onset              1
#>  8 2014-W18   f      infection          6
#>  9 2014-W18   f      onset              3
#> 10 2014-W19   f      infection          6
#> # ℹ 206 more rows

For a quick, high-level, overview of grouped data we can use the summary() method:

summary(weekly_multi_dates)
#> From:          2014-W15
#> To:            2015-W18
#> Groups:        gender
#> 
#> Total observations:
#> # A data frame: 4 × 3
#>   gender count_variable count
#>   <fct>  <chr>          <int>
#> 1 f      onset           2195
#> 2 m      infection       1390
#> 3 f      infection       1408
#> 4 m      onset           2178

When multiple date indices are given, they are used for rows of the resultant plot, unless the resultant variable is used to fill:

plot(weekly_multi_dates, angle = 45, border_colour = "white")

Four bar charts arranged in a 2 by 2 grid. The top row represents incidence by date of infection, the bottom row by date of onset. Each row is arranged with females in the left plots and males the right. The graphs all peak between 2014-W35 and 2014-W45. The "descent" from the peaks tapers off slower than the initial "ascent".

plot of chunk unnamed-chunk-12

plot(weekly_multi_dates, angle = 45, border_colour = "white", fill = "count_variable")

plot of chunk unnamed-chunk-13

3.2 Computing incidence from pre-aggregated data

In terms of this package, pre-aggregated data, is data where we have a single column representing time and associated counts linked to those times (still optionally split by characteristics). The included Covid data set is in this wide format with multiple count values given for each day.

covid <- subset(
    covidregionaldataUK,
    !region %in% c("England", "Scotland", "Northern Ireland", "Wales")
)
str(covid)
#> 'data.frame':	4410 obs. of  13 variables:
#>  $ date           : Date, format: "2020-01-30" "2020-01-30" ...
#>  $ region         : chr  "East Midlands" "East of England" "London" "North East" ...
#>  $ region_code    : chr  "E12000004" "E12000006" "E12000007" "E12000001" ...
#>  $ cases_new      : num  NA NA NA NA NA NA NA NA 1 NA ...
#>  $ cases_total    : num  NA NA NA NA NA NA NA NA 1 NA ...
#>  $ deaths_new     : num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ deaths_total   : num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ recovered_new  : num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ recovered_total: num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ hosp_new       : num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ hosp_total     : num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ tested_new     : num  NA NA NA NA NA NA NA NA NA NA ...
#>  $ tested_total   : num  NA NA NA NA NA NA NA NA NA NA ...

Like with our linelist data, incidence() requires us to specify a date_index column and optionally our groups and/or interval. In addition we must now also provide the counts variable(s) that we are interested in.

Before continuing, take note of the missing values in output above. Where these occur in one of the count variables, incidence() warns users:

monthly_covid <- incidence(
    covid,
    date_index = "date",
    groups = "region",
    counts = "cases_new",
    interval = "yearmonth"
)
#> Warning in incidence(): `cases_new` contains NA values. Consider imputing these
#> and calling `incidence()` again.
monthly_covid
#> # incidence:  162 x 4
#> # count vars: cases_new
#> # groups:     region
#>    date_index region                   count_variable count
#>    <yrmon>    <chr>                    <fct>          <dbl>
#>  1 2020-Jan   East Midlands            cases_new         NA
#>  2 2020-Jan   East of England          cases_new         NA
#>  3 2020-Jan   London                   cases_new         NA
#>  4 2020-Jan   North East               cases_new         NA
#>  5 2020-Jan   North West               cases_new         NA
#>  6 2020-Jan   South East               cases_new         NA
#>  7 2020-Jan   South West               cases_new         NA
#>  8 2020-Jan   West Midlands            cases_new         NA
#>  9 2020-Jan   Yorkshire and The Humber cases_new          1
#> 10 2020-Feb   East Midlands            cases_new         NA
#> # ℹ 152 more rows

Whilst we could have let incidence() ignore missing values (equivalent to setting sum(..., na.rm=TRUE)), we prefer that users make an explicit choice on how these should (or should not) be imputed. For example, to treat missing values as zero counts we can simply replace them in the data prior to calling incidence():

(monthly_covid <-
     covid |>
     tidyr::replace_na(list(cases_new = 0)) |> 
     incidence(
         date_index = "date",
         groups = "region",
         counts = "cases_new",
         interval = "yearmonth"
     ))
#> # incidence:  162 x 4
#> # count vars: cases_new
#> # groups:     region
#>    date_index region                   count_variable count
#>    <yrmon>    <chr>                    <fct>          <dbl>
#>  1 2020-Jan   East Midlands            cases_new          0
#>  2 2020-Jan   East of England          cases_new          0
#>  3 2020-Jan   London                   cases_new          0
#>  4 2020-Jan   North East               cases_new          0
#>  5 2020-Jan   North West               cases_new          0
#>  6 2020-Jan   South East               cases_new          0
#>  7 2020-Jan   South West               cases_new          0
#>  8 2020-Jan   West Midlands            cases_new          0
#>  9 2020-Jan   Yorkshire and The Humber cases_new          1
#> 10 2020-Feb   East Midlands            cases_new          4
#> # ℹ 152 more rows
plot(monthly_covid, nrow = 3, angle = 45, border_colour = "white")

$Nine bar charts arranged in a 3 by 3 grid representing incidence new covid cases by month across nine English regions. Each plot goes from the start of 2020 to mid 2021. In each plot we see an increase in cases towards the end of 2020 and in to early 2021.$

plot of chunk unnamed-chunk-16

3.3 Plotting in style of European Programme for Intervention Epidemiology Training (EPIET)

For small datasets it is convention of EPIET to display individual cases as rectangles. We can do this by setting show_cases = TRUE in the call to plot() which will display each case as an individual square with a white border.

dat <- ebola[160:180, ]

(small <- incidence(
    dat,
    date_index = "date_of_onset",
    date_names_to = "date"
))
#> # incidence:  9 x 3
#> # count vars: date_of_onset
#>   date       count_variable count
#>   <date>     <chr>          <int>
#> 1 2014-07-08 date_of_onset      2
#> 2 2014-07-09 date_of_onset      2
#> 3 2014-07-10 date_of_onset      1
#> 4 2014-07-11 date_of_onset      2
#> 5 2014-07-12 date_of_onset      3
#> 6 2014-07-13 date_of_onset      4
#> 7 2014-07-14 date_of_onset      5
#> 8 2014-07-15 date_of_onset      1
#> 9 2014-07-16 date_of_onset      1
plot(small, show_cases = TRUE, angle = 45, n_breaks = 10)

Bar chart of daily incidence covering the period 2014-07-08 to 2014-07-16 inclusive. It shows 21 cases, with each case represented by an individual square.

plot of chunk unnamed-chunk-17

(small_gender <- incidence(
    dat,
    date_index = "date_of_onset",
    groups = "gender",
    date_names_to = "date"
)) 
#> # incidence:  13 x 4
#> # count vars: date_of_onset
#> # groups:     gender
#>    date       gender count_variable count
#>    <date>     <fct>  <chr>          <int>
#>  1 2014-07-08 m      date_of_onset      2
#>  2 2014-07-09 f      date_of_onset      1
#>  3 2014-07-09 m      date_of_onset      1
#>  4 2014-07-10 f      date_of_onset      1
#>  5 2014-07-11 f      date_of_onset      1
#>  6 2014-07-11 m      date_of_onset      1
#>  7 2014-07-12 f      date_of_onset      3
#>  8 2014-07-13 f      date_of_onset      3
#>  9 2014-07-13 m      date_of_onset      1
#> 10 2014-07-14 f      date_of_onset      1
#> 11 2014-07-14 m      date_of_onset      4
#> 12 2014-07-15 f      date_of_onset      1
#> 13 2014-07-16 f      date_of_onset      1
plot(small_gender, show_cases = TRUE, angle = 45, n_breaks = 10, fill = "gender")

plot of chunk unnamed-chunk-18

4. Support for tidy-select semantics

When working interactively it can feel a little onerous constantly having to quote inputs for column names. To alleviate this we include the functions incidence_() and regroup_() which both support tidy-select semantics in their column arguments (i.e. date_index, groups and counts).

For now we have chosen to distinguish the functions via the post-fix underscore and have a preference for the standard version for non-interactive (e.g. programmatic usage). This could change over time if users feel having two similar functions is confusing.

5. Working with incidence objects

On top of the incidence constructor function and the basic plotting, printing and summary we provide a number of other useful functions and integrations for working with incidence2 objects.

Note: The following sections utilise tidy-select semantics and hence use the post-fix version of the incidence function (incidence_())

5.1 `regroup()`

If you’ve created a grouped incidence object but now want to change the internal grouping, you can regroup() to the desired aggregation:

# generate an incidence object with 3 groups
(x <- incidence_(
    ebola,
    date_index = date_of_onset,
    groups = c(gender, hospital, outcome),
    interval = "isoweek"
))
#> # incidence:  1,167 x 6
#> # count vars: date_of_onset
#> # groups:     gender, hospital, outcome
#>    date_index gender hospital                       outcome count_variable count
#>    <isowk>    <fct>  <fct>                          <fct>   <chr>          <int>
#>  1 2014-W15   f      Military Hospital              <NA>    date_of_onset      1
#>  2 2014-W16   m      Connaught Hospital             <NA>    date_of_onset      1
#>  3 2014-W17   f      other                          Recover date_of_onset      2
#>  4 2014-W17   m      other                          Recover date_of_onset      1
#>  5 2014-W18   f      Connaught Hospital             Recover date_of_onset      1
#>  6 2014-W18   f      Princess Christian Maternity … Death   date_of_onset      1
#>  7 2014-W18   f      Rokupa Hospital                Recover date_of_onset      1
#>  8 2014-W19   f      Connaught Hospital             <NA>    date_of_onset      1
#>  9 2014-W19   f      Connaught Hospital             Recover date_of_onset      1
#> 10 2014-W19   f      Military Hospital              <NA>    date_of_onset      2
#> # ℹ 1,157 more rows
# regroup to just two groups
regroup_(x, c(gender, outcome))
#> # incidence:  314 x 5
#> # count vars: date_of_onset
#> # groups:     gender, outcome
#>    date_index gender outcome count_variable count
#>    <isowk>    <fct>  <fct>   <chr>          <int>
#>  1 2014-W15   f      <NA>    date_of_onset      1
#>  2 2014-W16   m      <NA>    date_of_onset      1
#>  3 2014-W17   f      Recover date_of_onset      2
#>  4 2014-W17   m      Recover date_of_onset      1
#>  5 2014-W18   f      Death   date_of_onset      1
#>  6 2014-W18   f      Recover date_of_onset      2
#>  7 2014-W19   f      <NA>    date_of_onset      4
#>  8 2014-W19   f      Death   date_of_onset      1
#>  9 2014-W19   f      Recover date_of_onset      3
#> 10 2014-W19   m      <NA>    date_of_onset      1
#> # ℹ 304 more rows
# standard (non-tidy-select) version
regroup(x, c("gender", "outcome"))
#> # incidence:  314 x 5
#> # count vars: date_of_onset
#> # groups:     gender, outcome
#>    date_index gender outcome count_variable count
#>    <isowk>    <fct>  <fct>   <chr>          <int>
#>  1 2014-W15   f      <NA>    date_of_onset      1
#>  2 2014-W16   m      <NA>    date_of_onset      1
#>  3 2014-W17   f      Recover date_of_onset      2
#>  4 2014-W17   m      Recover date_of_onset      1
#>  5 2014-W18   f      Death   date_of_onset      1
#>  6 2014-W18   f      Recover date_of_onset      2
#>  7 2014-W19   f      <NA>    date_of_onset      4
#>  8 2014-W19   f      Death   date_of_onset      1
#>  9 2014-W19   f      Recover date_of_onset      3
#> 10 2014-W19   m      <NA>    date_of_onset      1
#> # ℹ 304 more rows
# drop all groups
regroup(x)
#> # incidence:  56 x 3
#> # count vars: date_of_onset
#>    date_index count_variable count
#>    <isowk>    <chr>          <int>
#>  1 2014-W15   date_of_onset      1
#>  2 2014-W16   date_of_onset      1
#>  3 2014-W17   date_of_onset      3
#>  4 2014-W18   date_of_onset      3
#>  5 2014-W19   date_of_onset     11
#>  6 2014-W20   date_of_onset     13
#>  7 2014-W21   date_of_onset     11
#>  8 2014-W22   date_of_onset     15
#>  9 2014-W23   date_of_onset     13
#> 10 2014-W24   date_of_onset     13
#> # ℹ 46 more rows

5.2 `complete_dates()`

Sometimes your incidence data does not span consecutive units of time, or different groupings may cover different periods. To this end we provide a complete_dates() function which ensures a complete and identical range of dates are given counts (by default filling with a 0 value).

dat <- data.frame(
    dates = as.Date(c("2020-01-01", "2020-01-04")),
    gender = c("male", "female")
)

(incidence <- incidence_(dat, date_index = dates, groups = gender))
#> # incidence:  2 x 4
#> # count vars: dates
#> # groups:     gender
#>   date_index gender count_variable count
#>   <date>     <chr>  <chr>          <int>
#> 1 2020-01-01 male   dates              1
#> 2 2020-01-04 female dates              1
complete_dates(incidence)
#> # incidence:  8 x 4
#> # count vars: dates
#> # groups:     gender
#>   date_index gender count_variable count
#>   <date>     <chr>  <chr>          <int>
#> 1 2020-01-01 female dates              0
#> 2 2020-01-01 male   dates              1
#> 3 2020-01-02 female dates              0
#> 4 2020-01-02 male   dates              0
#> 5 2020-01-03 female dates              0
#> 6 2020-01-03 male   dates              0
#> 7 2020-01-04 female dates              1
#> 8 2020-01-04 male   dates              0

5.3 `keep_first()`, `keep_last()` and `keep_peaks()`

Once your data is grouped by date, you can select the first or last few entries based on a particular date grouping using keep_first() and keep_last():

weekly_incidence <- incidence_(
    ebola,
    date_index = date_of_onset,
    groups = hospital,
    interval = "isoweek"
)

keep_first(weekly_incidence, 3)
#> # incidence:  15 x 4
#> # count vars: date_of_onset
#> # groups:     hospital
#>    date_index hospital                                     count_variable count
#>    <isowk>    <fct>                                        <chr>          <int>
#>  1 2014-W15   Connaught Hospital                           date_of_onset      0
#>  2 2014-W16   Connaught Hospital                           date_of_onset      1
#>  3 2014-W17   Connaught Hospital                           date_of_onset      0
#>  4 2014-W15   Military Hospital                            date_of_onset      1
#>  5 2014-W16   Military Hospital                            date_of_onset      0
#>  6 2014-W17   Military Hospital                            date_of_onset      0
#>  7 2014-W15   other                                        date_of_onset      0
#>  8 2014-W16   other                                        date_of_onset      0
#>  9 2014-W17   other                                        date_of_onset      3
#> 10 2014-W15   Princess Christian Maternity Hospital (PCMH) date_of_onset      0
#> 11 2014-W16   Princess Christian Maternity Hospital (PCMH) date_of_onset      0
#> 12 2014-W17   Princess Christian Maternity Hospital (PCMH) date_of_onset      0
#> 13 2014-W15   Rokupa Hospital                              date_of_onset      0
#> 14 2014-W16   Rokupa Hospital                              date_of_onset      0
#> 15 2014-W17   Rokupa Hospital                              date_of_onset      0
keep_last(weekly_incidence, 3)
#> # incidence:  15 x 4
#> # count vars: date_of_onset
#> # groups:     hospital
#>    date_index hospital                                     count_variable count
#>    <isowk>    <fct>                                        <chr>          <int>
#>  1 2015-W16   Connaught Hospital                           date_of_onset     14
#>  2 2015-W17   Connaught Hospital                           date_of_onset     10
#>  3 2015-W18   Connaught Hospital                           date_of_onset      9
#>  4 2015-W16   Military Hospital                            date_of_onset      3
#>  5 2015-W17   Military Hospital                            date_of_onset      6
#>  6 2015-W18   Military Hospital                            date_of_onset      2
#>  7 2015-W16   other                                        date_of_onset      9
#>  8 2015-W17   other                                        date_of_onset      3
#>  9 2015-W18   other                                        date_of_onset      2
#> 10 2015-W16   Princess Christian Maternity Hospital (PCMH) date_of_onset      1
#> 11 2015-W17   Princess Christian Maternity Hospital (PCMH) date_of_onset      3
#> 12 2015-W18   Princess Christian Maternity Hospital (PCMH) date_of_onset      2
#> 13 2015-W16   Rokupa Hospital                              date_of_onset      3
#> 14 2015-W17   Rokupa Hospital                              date_of_onset      6
#> 15 2015-W18   Rokupa Hospital                              date_of_onset      2

Similarly keep_peaks()returns the rows corresponding to the maximum count value for each grouping of an incidence2 object:

keep_peaks(weekly_incidence)
#> # incidence:  5 x 4
#> # count vars: date_of_onset
#> # groups:     hospital
#>   date_index hospital                                     count_variable count
#>   <isowk>    <fct>                                        <chr>          <int>
#> 1 2014-W39   Connaught Hospital                           date_of_onset    100
#> 2 2014-W38   Military Hospital                            date_of_onset     54
#> 3 2014-W42   other                                        date_of_onset     52
#> 4 2014-W38   Princess Christian Maternity Hospital (PCMH) date_of_onset     25
#> 5 2014-W42   Rokupa Hospital                              date_of_onset     28

5.4 Bootstrapping and estimating peaks

estimate_peak() returns an estimate of the peak of an epidemic curve using bootstrapped samples of the available data. It is a wrapper around two functions:

Firstly, the imaginatively named, first_peak(), that returns the earliest occurring peak row per group; and,
Secondly, bootstrap_incidence() which samples (with replacement and optional randomisation) from incidence2 objects.

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

the total number of event is known (no uncertainty on total incidence);
dates with no events (zero incidence) will never be in bootstrapped datasets; and
the reporting is assumed to be constant over time, i.e. every case is equally likely to be reported.

influenza <- incidence_(
    outbreaks::fluH7N9_china_2013,
    date_index = date_of_onset,
    groups = province
)

# across provinces (we suspend progress bar for markdown)
estimate_peak(influenza, progress = FALSE) |> 
    select(-count_variable)
#> # A tibble: 13 × 7
#>    province  observed_peak observed_count bootstrap_peaks lower_ci   median    
#>    <fct>     <date>                 <int> <list>          <date>     <date>    
#>  1 Anhui     2013-03-09                 1 <df [100 × 1]>  2013-03-09 2013-03-28
#>  2 Beijing   2013-04-11                 1 <df [100 × 1]>  2013-02-19 2013-04-11
#>  3 Fujian    2013-04-17                 1 <df [100 × 1]>  2013-04-17 2013-04-18
#>  4 Guangdong 2013-07-27                 1 <df [100 × 1]>  2013-02-19 2013-07-27
#>  5 Hebei     2013-07-10                 1 <df [100 × 1]>  2013-02-19 2013-07-10
#>  6 Henan     2013-04-06                 1 <df [100 × 1]>  2013-04-06 2013-04-08
#>  7 Hunan     2013-04-14                 1 <df [100 × 1]>  2013-02-19 2013-04-14
#>  8 Jiangsu   2013-03-19                 2 <df [100 × 1]>  2013-03-08 2013-03-20
#>  9 Jiangxi   2013-04-15                 1 <df [100 × 1]>  2013-04-15 2013-04-19
#> 10 Shandong  2013-04-16                 1 <df [100 × 1]>  2013-02-19 2013-04-16
#> 11 Shanghai  2013-04-01                 4 <df [100 × 1]>  2013-03-17 2013-04-01
#> 12 Taiwan    2013-04-12                 1 <df [100 × 1]>  2013-02-19 2013-04-12
#> 13 Zhejiang  2013-04-06                 5 <df [100 × 1]>  2013-03-29 2013-04-08
#> # ℹ 1 more variable: upper_ci <date>
# regrouping for overall peak
plot(regroup(influenza))

Bar chart of daily incidence covering the period March 2013 to August 2013 inclusive. The graph appears to peak around the start of April.

plot of chunk unnamed-chunk-23

estimate_peak(regroup(influenza), progress = FALSE) |> 
    select(-count_variable)
#> # A tibble: 1 × 6
#>   observed_peak observed_count bootstrap_peaks lower_ci   median     upper_ci  
#>   <date>                 <int> <list>          <date>     <date>     <date>    
#> 1 2013-04-03                 7 <df [100 × 1]>  2013-03-29 2013-04-06 2013-04-14
# return the first peak of the grouped and ungrouped data
first_peak(influenza)
#> # incidence:  13 x 4
#> # count vars: date_of_onset
#> # groups:     province
#>    date_index province  count_variable count
#>    <date>     <fct>     <chr>          <int>
#>  1 2013-03-09 Anhui     date_of_onset      1
#>  2 2013-04-11 Beijing   date_of_onset      1
#>  3 2013-04-17 Fujian    date_of_onset      1
#>  4 2013-07-27 Guangdong date_of_onset      1
#>  5 2013-07-10 Hebei     date_of_onset      1
#>  6 2013-04-06 Henan     date_of_onset      1
#>  7 2013-04-14 Hunan     date_of_onset      1
#>  8 2013-03-19 Jiangsu   date_of_onset      2
#>  9 2013-04-15 Jiangxi   date_of_onset      1
#> 10 2013-04-16 Shandong  date_of_onset      1
#> 11 2013-04-01 Shanghai  date_of_onset      4
#> 12 2013-04-12 Taiwan    date_of_onset      1
#> 13 2013-04-06 Zhejiang  date_of_onset      5
first_peak(regroup(influenza))
#> # 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
# bootstrap a single sample
bootstrap_incidence(influenza)
#> # incidence:  84 x 4
#> # count vars: date_of_onset
#> # groups:     province
#>    date_index province count_variable count
#>  * <date>     <fct>    <chr>          <int>
#>  1 2013-02-19 Shanghai date_of_onset      0
#>  2 2013-02-27 Shanghai date_of_onset      1
#>  3 2013-03-07 Zhejiang date_of_onset      1
#>  4 2013-03-08 Jiangsu  date_of_onset      0
#>  5 2013-03-09 Anhui    date_of_onset      1
#>  6 2013-03-13 Zhejiang date_of_onset      0
#>  7 2013-03-17 Shanghai date_of_onset      0
#>  8 2013-03-19 Jiangsu  date_of_onset      4
#>  9 2013-03-20 Jiangsu  date_of_onset      2
#> 10 2013-03-21 Jiangsu  date_of_onset      3
#> # ℹ 74 more rows

5.5 `cumulate()`

You can use cumulate() to easily generate cumulative incidences:

(y <- cumulate(weekly_incidence))
#> # incidence:  261 x 4
#> # count vars: date_of_onset
#> # groups:     hospital
#>    date_index hospital           count_variable cumulative_count
#>  * <isowk>    <fct>              <chr>                     <int>
#>  1 2014-W16   Connaught Hospital date_of_onset                 1
#>  2 2014-W18   Connaught Hospital date_of_onset                 2
#>  3 2014-W19   Connaught Hospital date_of_onset                 5
#>  4 2014-W20   Connaught Hospital date_of_onset                 7
#>  5 2014-W21   Connaught Hospital date_of_onset                12
#>  6 2014-W22   Connaught Hospital date_of_onset                16
#>  7 2014-W23   Connaught Hospital date_of_onset                22
#>  8 2014-W24   Connaught Hospital date_of_onset                28
#>  9 2014-W25   Connaught Hospital date_of_onset                40
#> 10 2014-W26   Connaught Hospital date_of_onset                48
#> # ℹ 251 more rows
plot(y, angle = 45, nrow = 3)

Fives graphs representing cumulative weekly incidence covering 2014-W15 to 2015-W18 inclusive. Each graph represents a hospital in the data set. The five graphs fill a 3 by 2 grid with the bottom-right square being left blank.

plot of chunk unnamed-chunk-24

6. Building on incidence2

The benefit incidence2 brings is not in the functionality it provides (which is predominantly wrapping around the functionality of other packages) but in the guarantees incidence2 objects give to a user about the underlying object structure and invariants that must hold.

To make these objects easier to build upon we give sensible behaviour when the invariants are broken, an interface to access the variables underlying the incidence2 object and methods, for popular group-aware generics, that implicitly utilise the underlying grouping structure.

6.1 Class preservation

As mentioned at the beginning of the vignetted, by definition, incidence2 objects must:

have one column representing the date index (this does not need to be a Date object but must have an inherent ordering over time);
have one column representing the count variable (i.e. what is being counted) and one variable representing the associated count;
have zero or more columns representing groups;
not have duplicated rows with regards to the date, group and count variables.

Due to these requirements it is important that these objects preserve (or drop) their structure appropriately under the range of different operations that can be applied to data frames. By this we mean that if an operation is applied to an incidence2 object then as long as the invariants of the object are preserved (i.e. required columns and uniqueness of rows) then the object will retain it’s incidence class. If the invariants are not preserved then a tibble will be returned instead.

# create a weekly incidence object
weekly_incidence <- incidence_(
    ebola,
    date_index = date_of_onset,
    groups = c(gender, hospital),
    interval = "isoweek"
)

# filtering preserves class
weekly_incidence |> 
    subset(gender == "f" & hospital == "Rokupa Hospital") |> 
    class()
#> [1] "incidence2" "tbl_df"     "tbl"        "data.frame"

class(weekly_incidence[c(1L, 3L, 5L), ])
#> [1] "incidence2" "tbl_df"     "tbl"        "data.frame"

# Adding columns preserve class
weekly_incidence$future <- weekly_incidence$date_index + 999L
class(weekly_incidence)
#> [1] "incidence2" "tbl_df"     "tbl"        "data.frame"
weekly_incidence |> 
    mutate(past = date_index - 999L) |> 
    class()
#> [1] "incidence2" "tbl_df"     "tbl"        "data.frame"

# rename preserve class
names(weekly_incidence)[names(weekly_incidence) == "date_index"] <- "isoweek"
str(weekly_incidence)
#> incidnc2 [496 × 6] (S3: incidence2/tbl_df/tbl/data.frame)
#>  $ isoweek       : 'grates_isoweek' int [1:496] 2014-W15 2014-W16 2014-W17 2014-W17 2014-W18 2014-W18 2014-W18 2014-W19 2014-W19 2014-W19 ...
#>  $ gender        : Factor w/ 2 levels "f","m": 1 2 1 2 1 1 1 1 1 1 ...
#>  $ hospital      : Factor w/ 5 levels "Connaught Hospital",..: 2 1 3 3 1 4 5 1 2 4 ...
#>  $ count_variable: chr [1:496] "date_of_onset" "date_of_onset" "date_of_onset" "date_of_onset" ...
#>  $ count         : int [1:496] 1 1 2 1 1 1 1 2 5 1 ...
#>  $ future        : 'grates_isoweek' int [1:496] 2033-W22 2033-W23 2033-W24 2033-W24 2033-W25 2033-W25 2033-W25 2033-W26 2033-W26 2033-W26 ...
#>  - attr(*, "date_index")= chr "isoweek"
#>  - attr(*, "count_variable")= chr "count_variable"
#>  - attr(*, "count_value")= chr "count"
#>  - attr(*, "groups")= chr [1:2] "gender" "hospital"

# select returns a tibble unless all date, count and group variables are
# preserved in the output
str(weekly_incidence[,-1L])
#> tibble [496 × 5] (S3: tbl_df/tbl/data.frame)
#>  $ gender        : Factor w/ 2 levels "f","m": 1 2 1 2 1 1 1 1 1 1 ...
#>  $ hospital      : Factor w/ 5 levels "Connaught Hospital",..: 2 1 3 3 1 4 5 1 2 4 ...
#>  $ count_variable: chr [1:496] "date_of_onset" "date_of_onset" "date_of_onset" "date_of_onset" ...
#>  $ count         : int [1:496] 1 1 2 1 1 1 1 2 5 1 ...
#>  $ future        : 'grates_isoweek' int [1:496] 2033-W22 2033-W23 2033-W24 2033-W24 2033-W25 2033-W25 2033-W25 2033-W26 2033-W26 2033-W26 ...
str(weekly_incidence[, -6L])
#> incidnc2 [496 × 5] (S3: incidence2/tbl_df/tbl/data.frame)
#>  $ isoweek       : 'grates_isoweek' int [1:496] 2014-W15 2014-W16 2014-W17 2014-W17 2014-W18 2014-W18 2014-W18 2014-W19 2014-W19 2014-W19 ...
#>  $ gender        : Factor w/ 2 levels "f","m": 1 2 1 2 1 1 1 1 1 1 ...
#>  $ hospital      : Factor w/ 5 levels "Connaught Hospital",..: 2 1 3 3 1 4 5 1 2 4 ...
#>  $ count_variable: chr [1:496] "date_of_onset" "date_of_onset" "date_of_onset" "date_of_onset" ...
#>  $ count         : int [1:496] 1 1 2 1 1 1 1 2 5 1 ...
#>  - attr(*, "date_index")= chr "isoweek"
#>  - attr(*, "count_variable")= chr "count_variable"
#>  - attr(*, "count_value")= chr "count"
#>  - attr(*, "groups")= chr [1:2] "gender" "hospital"

# duplicating rows will drop the class but only if duplicate rows
class(rbind(weekly_incidence, weekly_incidence))
#> [1] "tbl_df"     "tbl"        "data.frame"
class(rbind(weekly_incidence[1:5, ], weekly_incidence[6:10, ]))
#> [1] "incidence2" "tbl_df"     "tbl"        "data.frame"

6.2 Accessing variable information

We provide multiple accessors to easily access information about an incidence2 object’s structure:

# the name of the date_index variable of x
get_date_index_name(weekly_incidence)
#> [1] "isoweek"
# alias for `get_date_index_name()`
get_dates_name(weekly_incidence)
#> [1] "isoweek"
# the name of the count variable of x
get_count_variable_name(weekly_incidence)
#> [1] "count_variable"
# the name of the count value of x
get_count_value_name(weekly_incidence)
#> [1] "count"
# the name(s) of the group variable(s) of x
get_group_names(weekly_incidence)
#> [1] "gender"   "hospital"
# the date_index variable of x
str(get_date_index(weekly_incidence))
#>  'grates_isoweek' int [1:496] 2014-W15 2014-W16 2014-W17 2014-W17 2014-W18 2014-W18 2014-W18 2014-W19 2014-W19 2014-W19 ...
# alias for get_date_index
str(get_dates(weekly_incidence))
#>  'grates_isoweek' int [1:496] 2014-W15 2014-W16 2014-W17 2014-W17 2014-W18 2014-W18 2014-W18 2014-W19 2014-W19 2014-W19 ...
# the count variable of x
str(get_count_variable(weekly_incidence))
#>  chr [1:496] "date_of_onset" "date_of_onset" "date_of_onset" ...
# the count value of x
str(get_count_value(weekly_incidence))
#>  int [1:496] 1 1 2 1 1 1 1 2 5 1 ...
# list of the group variable(s) of x
str(get_groups(weekly_incidence))
#> List of 2
#>  $ gender  : Factor w/ 2 levels "f","m": 1 2 1 2 1 1 1 1 1 1 ...
#>  $ hospital: Factor w/ 5 levels "Connaught Hospital",..: 2 1 3 3 1 4 5 1 2 4 ...

6.3 Grouping aware methods

incidence2 provides methods for popular group-aware generics from both base R and the wider package ecosystem:

base: split().
dplyr: mutate() and summarise().
tidyr: nest().

When called on incidence2 objects, these methods will utilise the underlying grouping structure without the user needing to explicitly state what it is. This makes it very easy to utilise in analysis pipelines.

6.3.1 Example fitting a poisson model

# first twenty weeks of the ebola data set across hospitals
dat <- incidence_(ebola, date_of_onset, groups = hospital, interval = "isoweek")
dat <- keep_first(dat, 20L)

# fit a poisson model to the grouped data
(fitted <-
    dat |>
    nest(.key = "data") |>
    mutate(
        model  = lapply(
            data,
            function(x) glm(count ~ date_index, data = x, family = poisson)
        )
    ))
#> # A tibble: 5 × 4
#>   count_variable hospital                                     data     model 
#>   <chr>          <fct>                                        <list>   <list>
#> 1 date_of_onset  Connaught Hospital                           <tibble> <glm> 
#> 2 date_of_onset  Military Hospital                            <tibble> <glm> 
#> 3 date_of_onset  other                                        <tibble> <glm> 
#> 4 date_of_onset  Princess Christian Maternity Hospital (PCMH) <tibble> <glm> 
#> 5 date_of_onset  Rokupa Hospital                              <tibble> <glm>
# Add confidence intervals to the result
(intervals <-
    fitted |>
    mutate(result = Map(
        function(data, model) {
            data |>
                ciTools::add_ci(
                    model,
                    alpha = 0.05,
                    names = c("lower_ci", "upper_ci")
                ) |>
                as_tibble()
        },
        data,
        model
    )) |>
    select(hospital, result) |>
    unnest(result))
#> # A tibble: 100 × 6
#>    hospital           date_index count  pred lower_ci upper_ci
#>    <fct>              <isowk>    <int> <dbl>    <dbl>    <dbl>
#>  1 Connaught Hospital 2014-W15       0 0.968    0.599     1.56
#>  2 Connaught Hospital 2014-W16       1 1.18     0.751     1.85
#>  3 Connaught Hospital 2014-W17       0 1.44     0.942     2.19
#>  4 Connaught Hospital 2014-W18       1 1.75     1.18      2.58
#>  5 Connaught Hospital 2014-W19       3 2.13     1.48      3.06
#>  6 Connaught Hospital 2014-W20       2 2.59     1.86      3.61
#>  7 Connaught Hospital 2014-W21       5 3.16     2.33      4.28
#>  8 Connaught Hospital 2014-W22       4 3.84     2.91      5.07
#>  9 Connaught Hospital 2014-W23       6 4.68     3.65      6.00
#> 10 Connaught Hospital 2014-W24       6 5.70     4.56      7.12
#> # ℹ 90 more rows
# plot
plot(dat, angle = 45) +
    ggplot2::geom_line(
        ggplot2::aes(date_index, y = pred),
        data = intervals,
        inherit.aes = FALSE
    ) +
    ggplot2::geom_ribbon(
        ggplot2::aes(date_index, ymin = lower_ci, ymax = upper_ci),
        alpha = 0.2,
        data = intervals,
        inherit.aes = FALSE,
        fill = "#BBB67E"
    )

Fives bar charts representing weekly incidence covering 2014-W15 to 2014-W35 inclusive. Each graph represents a hospital in the data set. The five graphs fill a 3 by 2 grid with the bottom-right square being left blank. On top of each bar graph is a line along with associated confidence intervals showing an increasing trend over the displayed weeks.

plot of chunk unnamed-chunk-27

6.3.2 Example - Adding a rolling average

weekly_incidence |>
    regroup_(hospital) |> 
    mutate(rolling_average = data.table::frollmean(count, n = 3L, align = "right")) |> 
    plot(border_colour = "white", angle = 45) +
    ggplot2::geom_line(ggplot2::aes(x = date_index, y = rolling_average))
#> Warning: Removed 2 rows containing missing values or values outside the scale range
#> (`geom_line()`).

Fives bar charts representing weekly incidence covering 2014-W15 to 2015-W18 inclusive. Each graph represents a hospital in the data set. The five graphs fill a 3 by 2 grid with the bottom-right square being left blank. On top of each bar graph is a line along that displays the rolling average.

plot of chunk unnamed-chunk-28

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.