Dummy Encoding

For some ML methods it is helpful to transform a column with categories into multiple categories with binary values.

For instance, we have a column crop that contains the values barley, wheat and rice.

With dummy encoding, we create three columns crop_barley,crop_wheat and crop_rice, each having values 0 or 1.

crops <- c("wheat", "rice", "rice", "wheat", "barley", "rice", "rice")

sizes <- c(3, 5, 6, 3, 4, 2, 2)

# format as data frame

data <- data.frame(crop=crops, size=sizes)

table(data)

# dummy encoding

dummies <- model.matrix(~crops-1, data=data)

dummies

data$crops = dummies