create_dummy_variables: Simple function to create dummy variables for ordinal and...

In mfp2: Multivariable Fractional Polynomial Models with Extensions

Simple function to create dummy variables for ordinal and nominal variables

Description

Simple function to create dummy variables for ordinal and nominal variables

Usage

create_dummy_variables(
  data,
  var_ordinal = NULL,
  var_nominal = NULL,
  drop_variables = FALSE
)

Arguments

data	A dataframe containing the ordinal variable.
var_ordinal	Names of ordinal variables in the data for which dummy variables should be created.
var_nominal	Names of nominal variables in the data for which dummy variables should be created.
drop_variables	Specifies whether to drop the original variables after dummy variables have been created. The default value is FALSE, and the original variables are kept in the data.

Details

This function creates dummy variables based on ordinal and categorical coding described in the Royston and Sauerbrei (2008) book (Chapter 3, Table 3.5). It uses the levels of the categorical variable if they exist; otherwise, it will extract the unique values of the variable, sort them, and use them as levels. We recommend that the user sets the levels of categorical variables and specifies their reference group. You can use the factor() function in base R. If the levels are 1, 2, and 3, then 1 will be the reference group. On the other hand, if the levels are 3, 2, and 1, then 3 will be the reference group. In brief, the first level will be taken as the reference group.

Value

A dataframe with new dummy variables.

Examples

data("gbsg")
# create dummy variable for grade using ordinal coding
gbsg <- create_dummy_variables(gbsg, var_ordinal = "grade", drop_variables = TRUE)
head(gbsg)

mfp2 documentation built on Nov. 15, 2023, 1:06 a.m.