Skip to contents

Identify outliers in a numeric vector

Usage

utilMWRoutlier(x, logscl)

Arguments

x

numeric vector of any length

logscl

logical to indicate if vector should be log10-transformed first

Value

A logical vector equal in length to x indicating TRUE for outliers or FALSE for within normal range

Details

Outliers are identified as 1.5 times the interquartile range

Examples

x <- rnorm(20)
utilMWRoutlier(x, logscl = FALSE)
#>  [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [13] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE