remove.zero.var.factors: Identify, alert, and remove columns with near zero variance.

In dataframing/archive: Archive: A personal collection of utilities

Description

This function is yet another fairly thin wrapper around nearZeroVar. caret's function returns a vector of indices that are near zero variance, and this function goes further to (a) inform the analyst of what columns are being removed and (b) removing those columns.

Usage

remove.zero.var.factors(data)

Arguments

data	any dataframe-esque object containing a mix of factor and numerical columns.

Examples

x <- data.frame(
  id     = seq(1, 10),
  age    = rpois(n=10, lambda=20),
  gender = rep(c('M', 'F'), 5),
  blood.type = factor(rep(c('A', 'B'), 5)),
  debt = factor(1e6)
)

# Removes only `debt` because (1) it's a factor and (2) has zero variance.
remove.zero.var.factors(x)

dataframing/archive documentation built on May 20, 2019, 10:19 p.m.