impact_code: transform high cardinality categorical variables

In mattmills49/modeler: Exploratory Data Analysis and Modeling Helper Functions

Description

impact_code transforms a categorical variable into a continuous one using Impact Coding described in this paper.

Usage

impact_code(.data, formula, binary = T)

Arguments

.data	the data frame the data is contained in
formula	a formula specifying the dependent and independent variables to use
binary	is the dependent variable a binary variable? Logical value

Details

Missing values in the categorical variable will be treated as another level. If you want to ignore missing values you will need to do so prior to calling impact_code.

Value

a data frame that includes the levels of the categorical variable and the smoothed estimate of the dependent variable.

Examples

impact_code(mtcars, am ~ cyl)
impact_code(iris, Petal.Width ~ Species, binary = F)

mattmills49/modeler documentation built on May 21, 2019, 1:25 p.m.