reducelevels: Reducing the amount of levels in factor columns

In DavidArenburg/dt.nuggets: Some convinience functions using the data.table package

Description

Reducing the amount of levels in factor columns

Usage

reducelevels(DT, cols, freq = 0.1, newlvl = "OTHER")

Arguments

DT	The data.table to operate on
cols	Quoted column names
freq	The quantile from lower eqaul to you want to replace levels
newlvl	The name of the new level

Value

The modified data.table with the refactored columns

Examples

set.seed(123)
DT <- data.table(A = factor(sample(letters, 1e3L, replace = TRUE, prob = (1:length(letters))/100)),
                 B = factor(sample(letters, 1e3L, replace = TRUE, prob = rev(1:length(letters))/100)),
                 C = factor(sample(letters, 1e3L, replace = TRUE, prob = sample(1:length(letters))/100)),
                 D = sample(letters, 1e3, replace = TRUE),
                 E = sample(1e3))

str(DT)
reducelevels(DT)
str(DT)
reducelevels(DT, c("A", "B"), 0.6)
str(DT)

DavidArenburg/dt.nuggets documentation built on May 6, 2019, 1:54 p.m.