Summarise categorical measure(s) by group(s)

For each specified grouping variable, count the frequency of each unique category of the given measure column(s), and compute the denominator (non-missing categories by default) and percentage. Results for the specified group(s) are combined into a single long tibble for easy use in `ggplot` or `plotly`.

Usage

get_frequency(
  data,
  measures,
  groups = "overall",
  nested = FALSE,
  count_na = FALSE
)

Arguments

data: A data frame containing measure columns (and grouping columns), with one row per participant.
measures: A character vector of column names to summarise. All columns will be coerced to character, so this works for logical, factor, and character columns alike. Measures appear in the output in this order.
groups: A character vector of grouping column names. Defaults to "overall", which creates a single group containing all rows. Rows where the grouping variable is NA are excluded from that group's summary.
nested: Logical; if FALSE (default), each grouping variable in groups is summarised separately. If TRUE, all variables in groups are treated as a nested grouping set.
count_na: Logical; if FALSE (default), missing values in measures are excluded from the numerator and denominator. If TRUE, missing values are counted as their own category (category = NA) and included in the denominator, but only for measures that contain at least one missing value (like table(useNA = "ifany")). Missing values in groups are always excluded.

Value

A tibble in long format with one row per group level × measure × category combination.

Examples


library(dplyr)
#> 
#> Attaching package: ‘dplyr’
#> The following objects are masked from ‘package:stats’:
#> 
#>     filter, lag
#> The following objects are masked from ‘package:base’:
#> 
#>     intersect, setdiff, setequal, union

set.seed(1999)
df <- data.frame(
  participant_id = 1:60,
  country        = c(rep("England", 30), rep("Wales", 30)),
  region         = c(
    rep("East England", 10), rep("West England", 10), rep(NA, 10),
    rep("North Wales", 10), rep("South Wales", 10), rep(NA, 10)
  ),
  # Categorical Q1
  q1_catq        = sample(c("A", "B", "C", NA), 60, replace = TRUE),
  # Categorical Q2
  q2_catq        = sample(c("A", "B", "C", "D", "E", NA), 60, replace = TRUE)
)

group_cols <- c("overall", "country", "region")

measure_cols <- df |>
  select(tidyselect::matches("q[0-9]+_catq")) |>
  names()

freq <- get_frequency(
  data     = df,
  measures = measure_cols,
  groups   = group_cols
)

head(freq)
#> # A tibble: 6 × 8
#>   overall country region measure category numerator denominator percent
#>   <chr>   <chr>   <chr>  <chr>   <chr>        <dbl>       <dbl>   <dbl>
#> 1 overall NA      NA     q1_catq A               10          41   0.244
#> 2 overall NA      NA     q1_catq B               18          41   0.439
#> 3 overall NA      NA     q1_catq C               13          41   0.317
#> 4 overall NA      NA     q2_catq A                9          54   0.167
#> 5 overall NA      NA     q2_catq B               16          54   0.296
#> 6 overall NA      NA     q2_catq C               10          54   0.185

# Nested grouping example (region is nested within country):
sum_q1_nested <- get_frequency(
  data = df,
  measures = "q1_catq",
  groups = c("country", "region"),
  nested = TRUE
)

sum_q1_nested
#> # A tibble: 12 × 7
#>    country region       measure category numerator denominator percent
#>    <chr>   <chr>        <chr>   <chr>        <int>       <int>   <dbl>
#>  1 England East England q1_catq A                2           7   0.286
#>  2 England East England q1_catq B                4           7   0.571
#>  3 England East England q1_catq C                1           7   0.143
#>  4 England West England q1_catq A                1           7   0.143
#>  5 England West England q1_catq B                3           7   0.429
#>  6 England West England q1_catq C                3           7   0.429
#>  7 Wales   North Wales  q1_catq A                1           7   0.143
#>  8 Wales   North Wales  q1_catq B                2           7   0.286
#>  9 Wales   North Wales  q1_catq C                4           7   0.571
#> 10 Wales   South Wales  q1_catq A                3           9   0.333
#> 11 Wales   South Wales  q1_catq B                4           9   0.444
#> 12 Wales   South Wales  q1_catq C                2           9   0.222

# Count missing values as a category
get_frequency(
  data = df,
  measures = "q1_catq",
  groups = "country",
  count_na = TRUE
)
#> # A tibble: 8 × 6
#>   country measure category numerator denominator percent
#>   <chr>   <chr>   <chr>        <int>       <int>   <dbl>
#> 1 England q1_catq A                5          30   0.167
#> 2 England q1_catq B               10          30   0.333
#> 3 England q1_catq C                6          30   0.2  
#> 4 England q1_catq NA               9          30   0.3  
#> 5 Wales   q1_catq A                5          30   0.167
#> 6 Wales   q1_catq B                8          30   0.267
#> 7 Wales   q1_catq C                7          30   0.233
#> 8 Wales   q1_catq NA              10          30   0.333