count_checkboxes: Count checkbox responses

In emilelatour/redcaptools: Misc Tools For Working With REDCap Data In R

Count checkbox responses

Description

Get summary counts easily from checkbox style variables/columns in a REDCap data set.

Usage

count_checkboxes(
  data,
  ...,
  check_text = "Checked",
  negate = FALSE,
  key_text = "field_name",
  value_text = "n",
  data_labels = NULL
)

Arguments

data	A tbl_df or data.frame
...	Checkbox column names. Supports 'dplyr' 'select_helpers'
check_text	Character string of the response to count. Default is "Checked"
negate	Boolean. If 'TRUE', the count is '!= check_text'. Default is 'FALSE'
key_text	Character string to name result column of the columns
value_text	Character string to name the result column of counts
data_labels	A data frame or tibble with columns, 'field_name' and 'checkbox_choice'. Used to replace variable names with checkbox labels.

Value

a tble_df

Examples

#### Example 1 --------------------------------

library(dplyr)
library(tibble)
library(tidyr)

df <- tibble::tibble(
  id = c(1:100),
  group = sample(c("A", "B", "C"), size = 100, replace = TRUE),
  preferred_color___1 = sample(c("Checked", "Unchecked"), size = 100, replace = TRUE),
  preferred_color___2 = sample(c("Checked", "Unchecked"), size = 100, replace = TRUE),
  preferred_color___3 = sample(c("Checked", "Unchecked"), size = 100, replace = TRUE),
  preferred_color___4 = sample(c("Checked", "Unchecked"), size = 100, replace = TRUE),
  preferred_color___5 = sample(c("Checked", "Unchecked"), size = 100, replace = TRUE)) %>%
  mutate(group = factor(group,
                        levels = c("A", "B", "C")))


var_labels <- tibble::tribble(
  ~field_name, ~checkbox_choice,
  "preferred_color___1",           "Blue",
  "preferred_color___2",          "Green",
  "preferred_color___3",         "Orange",
  "preferred_color___4",            "Red",
  "preferred_color___5",         "Yellow"
)

var_labels

# Different ways of selecting variables
count_checkboxes(data = df,
                 preferred_color___1:preferred_color___5)

count_checkboxes(data = df,
                 preferred_color___1, preferred_color___5)

count_checkboxes(data = df,
                 dplyr::starts_with("preferred_color"))


# Different ways to select the "Unchecked" text instead
count_checkboxes(data = df,
                 preferred_color___1:preferred_color___5,
                 check_text = "Unchecked",
                 key_text = "checkbox",
                 value_text = "count")


df %>%
  count_checkboxes(data = .,
                   preferred_color___1:preferred_color___5,
                   negate = TRUE)

# Can handle grouping
df %>%
  group_by(group) %>%
  count_checkboxes(data = .,
                   preferred_color___1:preferred_color___5)

# Can also apply labels instead of variable names
count_checkboxes(data = df,
                 preferred_color___1:preferred_color___5,
                 negate = TRUE,
                 data_labels = var_labels)



#### Example 2 --------------------------------

library(dplyr)
library(tibble)
library(tidyr)

df2 <- tibble::tibble(
  record_id = c(1:100),
  gender = sample(c("Female", "Male"),
                  size = 100,
                  replace = TRUE),
  gender_other = NA,
  age = sample(c(18:85),
               size = 100,
               replace = TRUE),
  education = sample(c("High-school", "College", "Graduate school"),
                     size = 100,
                     replace = TRUE),
  ethnicity = sample(c("Hispanic", "Non-hispanic"),
                     size = 100,
                     replace = TRUE),
  key = sample(c("race___1", "race___2", "race___3",
                 "race___4", "race___5", "race___6", "race___98"),
               size = 100,
               replace = TRUE),
  value = "Checked",
  income = sample(c(20000:120000),
                  size = 100,
                  replace = TRUE),
  marital_status = sample(c("Married", "Single"),
                          size = 100,
                          replace = TRUE),
  survey_complete = sample(c("Complete", "Not complete"),
                           size = 100,
                           replace = TRUE)) %>%
  mutate(key = factor(key,
                      levels = c("race___1", "race___2", "race___3",
                                 "race___4", "race___5", "race___6",
                                 "race___98"))) %>%
  tidyr::spread(.,
                key = "key",
                value = "value") %>%
  mutate_at(.vars = vars(race___1:race___98),
            .funs = list(~ tidyr::replace_na(., "Unchecked")))


data_labels <- tibble::tribble(
        ~field_name,                      ~checkbox_choice,
        "record_id",                                    NA,
           "gender",                                    NA,
     "gender_other",                                    NA,
              "age",                                    NA,
        "education",                                    NA,
        "ethnicity",                                    NA,
         "race___1",                               "White",
         "race___2",           "Black or African-American",
         "race___3",    "American Indian or Alaska Native",
         "race___4",                               "Asian",
         "race___5", "Native Hawaiian or Pacific Islander",
         "race___6",                               "Other",
        "race___98",                "Prefer not to answer",
       "race_other",                                    NA,
           "income",                                    NA,
   "marital_status",                                    NA,
  "survey_complete",                                    NA
  )




count_checkboxes(data = df2,
                 race___1:race___98)

count_checkboxes(data = df2,
                 dplyr::starts_with("race"))

count_checkboxes(data = df2,
                 dplyr::starts_with("race___"))

df2 %>%
  group_by(gender) %>%
  count_checkboxes(data = .,
                   race___1:race___98)

df2 %>%
  group_by(gender, education) %>%
  count_checkboxes(data = .,
                   race___1:race___98)



# With labels

count_checkboxes(data = df2,
                 race___1:race___98,
                 data_labels = data_labels)

df2 %>%
  group_by(gender, education) %>%
  count_checkboxes(data = .,
                   race___1:race___98,
                   data_labels = data_labels)

emilelatour/redcaptools documentation built on March 21, 2023, 3:35 p.m.