calcDepth_asymp: Calculating the K-sign depth (asymptotically)
In melaniehorn/GSignTest: Robust Tests for Regression-Parameters via Sign Depth

Description Usage Arguments Value Note References Examples

View source: R/calcDepth_asymp.R

calcDepth_asymp calculates the K-sign depth of a given vector of residuals (asymptotically).

1	calcDepth_asymp(res, K, transform = FALSE, exact = TRUE)

`res`	[`numeric`] numeric vector of residuals
`K`	[`integer(1)`] the parameter K for the sign depth
`transform`	[`logical(1)`] Should the depth be transformed by the formula N (res - 0.5^(K-1))*, so that the distribution of the K-sign depth converges for large number of data points N? Default is `FALSE`.
`exact`	[`logical(1)`] Should the depth be calculated exactly? If `FALSE`, terms of scale o(1) will ne ignored. At the moment, for exact results this method is available only for K = 2, 3, 4, 5. Default is `TRUE`.

[numeric(1)] the calculated K-sign depth of res.

The implementation is based on a work of Dennis Malcherczyk.

Malcherczyk, D. (2021+). K-sign depth: Asymptotic distribution, efficient computation and applications. Dissertation in preparation.

1
2
3

calcDepth_asymp(rnorm(10), 3)
calcDepth_asymp(runif(100, -1, 1), 4, transform = TRUE)
calcDepth_asymp(runif(100, -1, 1), 4, transform = TRUE, exact = FALSE)