create_descriptive_table: Format descriptive statistics tables
In uo-cmor/formattr: Format Values Nicely for Printing

View source: R/descriptive-tables.R

create_descriptive_table

R Documentation

Format descriptive statistics tables

Description

Construct nicely formatted descriptive statistics tables (mean/sd, n/percent).

Usage

create_descriptive_table(
  df,
  continuous = NULL,
  discrete = NULL,
  multiresponse = NULL,
  output = c(names(continuous), names(discrete), names(multiresponse)),
  by = NULL,
  value_continuous = mean_sd(.mean, .sd, accuracy = 0.1),
  value_discrete = n_percent(.n, .proportion),
  total = !is.null(by)
)

Arguments

`df`	A data frame containing the variables to describe.
`continuous, discrete`	Character vectors of continuous and discrete variables for which to calculate descriptive statistics. Can be named, in which case the provided names will be used in place of variable names in the first column of the table.
`multiresponse`	List of character vectors, each specifying the individual variables making up a combined multiple-response variable. Should be named, with the names giving the variable names to be used for the combined variables.
`output`	(optional) Character vector of variables to include in the table (and their order). Elements should be the names of elements of `continuous`, `discrete`, and `multireponse`. Default is to report all variables from `continuous`, `discrete`, and `multiresponse`, in that order.
`by`	(optional) A grouping variable (as a character string) to split the sample. Default is to report for the entire sample.
`value_continuous`	(optional) An expression describing how to format the outcome values for continuous variables. Can refer to mean/sd/min/lq/median/uq/max as `.mean`, `.sd`, etc, and the name of the variable being formatted as `.variable`. Default is to report mean (SD) to one decimal place.
`value_discrete`	(optional) An expression describing how to format the outcome values for discrete variables. Can refer to n/proportion as `.n`, `.proportion`, and the name of the variable being formatted as `.variable`. Default is to report n (percentage) to zero decimal places.
`total`	(optional) Logical value; should a 'total' (full sample) column be included? Default is `TRUE` if `by` is specified, `FALSE` otherwise. Can also be a character string giving the column name to use for the total column (default is 'Entire sample').