Skip to contents

Rousseeuw & Croux (1993) Qn robust scale

Usage

QnScale(y, constant = 2.21914, finite_corr = TRUE)

Arguments

y

numeric sample, length >= 2

constant

consistency constant, 1/(sqrt(2) qnorm(5/8)) = 2.21914

finite_corr

apply the robustbase finite-sample bias correction

Value

list(estimate, raw, k, h, n_pairs, correction, constant, n)

Examples

V <- c(1, 2, 3, 4, 5, 6, 7, 8)
QnScale(V)
#> $estimate
#> [1] 2.973337
#> 
#> $raw
#> [1] 2
#> 
#> $k
#> [1] 10
#> 
#> $h
#> [1] 5
#> 
#> $n_pairs
#> [1] 28
#> 
#> $correction
#> [1] 0.66993
#> 
#> $constant
#> [1] 2.21914
#> 
#> $n
#> [1] 8
#> 
#> $method
#> [1] "Rousseeuw & Croux (1993) Qn, robustbase qnsn.R definition"
#>