Downsample majority class to the size of minority class

Downsamples the majority class in binary classification setting to the size of minority class by randomly selecting majority class to match the number of minority class

downsample_majority(x, classVarName, arrangeVarName, seed = 100,
  shuffle = FALSE)

Arguments

x

A dataframe
classVarName

Name of the binary class column
arrangeVarName

Name of the variable to arrange by
seed

Seed chosen for sampling, defaults to 100
shuffle

Whether to shuffle the dataframe at the end. This cannot be TRUE when arrangeVarName is specified. Shuffle used the seed value seed + 1.

Examples

iris2         <- iris[c(1:25, 51:100), ]
iris2$Species <- factor(iris2$Species)
sampled <- downsample_majority(iris2, "Species")
table(sampled$Species)
#> 
#>     setosa versicolor 
#>         25         25