tabkris_2: Data description function

In CRUF: Clinical Research Utilities Functions

Description

tabkris_2 computes descriptive statistics for data

Usage

tabkris_2(
  data,
  names = NULL,
  varint = NULL,
  lang = "en",
  method = NULL,
  test = FALSE,
  pres_quant = c("med"),
  pres_quali = c("n", "per"),
  default_method = c("cont", "bino", "cate", "ordo"),
  default_test = c("stud", "chisq", "chisq", "chisq"),
  explicit_na = FALSE,
  digits = 2,
  return_table = TRUE,
  auto_detect = TRUE,
  lev_co = 10,
  verbose = FALSE
)

Arguments

data	Dataframe to describe or a "desctable" object
names	Vectors of variables to display in the final table, length of ncol(data)
varint	Variable to stratify on, factor only
lang	Language to display, default "en", "fr"
method	Vectors of variables to customize the methods used for description, length of data columns
test	Either a logical indicating statistical tests execution or a vectors of variables to customize the tests, length of data columns. Default FALSE
pres_quant	Descriptive statistics for quantitative variables. Possible values are "mean" for mean, SD, "med" for median, IQR, "range" for range
pres_quali	Descriptive statistics for qualitative variables. Possible values are "n" for number, "total" to add "/ total" and "per" for percentages
default_method	Default method to compute the table for each variable. Default default_method = c("cont", "bino", "cate", "ordo")
default_test	Default test to apply for each variable type. Default c("stud", "chisq", "chisq", "chisq"). Available "stud", "wilcox", "kruskal", "chisq", "fish"
explicit_na	Whether to display NA in description, Default FALSE
digits	Number of significant number to display, default 2
return_table	Whether to return a dataframe or an object to customize option easily, default TRUE
auto_detect	Whether to automatically detect variable type, transforming to factors numeric variable with moderate levels (< 10), default TRUE. Possible to set the cut-off number with lev_co
lev_co	Numeric. When auto_detect is TRUE, set the number of level to cutoff for categorical variables
verbose	Logical. Display information about transformation of variables. default FALSE

Details

The tabkris_2 function is a function to describe a set of data. Main purpose is to create a typical table one in biomedical litterature, either a patient characteristic table or population characteristic table.

names is a vector to name the variable of data. Default will use the colnames of data.

varint is a variable to stratify the analysis. It must be included in the initial dataset. It will not be displayed in the final table if chosen as the stratifying variable

lang is useful to choose the language for the final display. The default is english. French is also supported.

default_method and method are used to set the methods used for display. default_method must be length 4, to set the default method for continuous, binomial, categorical and ordered variable. method must be length of data columns, used to fine-tune every method for each variable.

default_test and test are used to set the tests performed. default_test must be lenght 4, to set the default method for continuous, binomial, categorical and ordered variable. test must be length of data columns, used to fine-tune every test for each variable.

pres_quant is used to set the display of quantitative variable. mean (SD), median [IQR] and range are available, default is median.

pres_quali is used to set the display of qualitative variable. "n" for number, "total" to add "/ total" and "per" for percentages, default is "n / per".

explicit_na is used to display.

digits is the number of digits to display for numbers. Usually if n < 100, digits = 0 if 100< n < 200, digits = 1 else digits = 2.

return_table choose if the user wants to directly display a table or if the user wants to get an object with parametrable objects.

auto_detect will test if each column can be coerced to a factor (i.e. having between 2 and 10 levels) and change the type of variable if so.

lev_co will set the number of maximum levels to coerce a column in a factor

Value

Depending on argument return_table, an object of class data.frame, which is the descriptive table or an object of class "desctable", which is a customizable object.

Author(s)

Yves Gallien [e-mail], 2019

See Also

https://github.com/Ygall/CRUF for manual and examples.

Examples

tabkris_2(boys)

CRUF documentation built on March 13, 2020, 1:34 a.m.