depth.qhpeeling: Calculate Convex Hull Peeling Depth
In ddalpha: Depth-Based Classification and Calculation of Data Depth

depth.qhpeeling

R Documentation

Calculate Convex Hull Peeling Depth

Description

Calculates the convex hull peeling depth of points w.r.t. a multivariate data set.

Usage

depth.qhpeeling(x, data)

Arguments

`x`	Matrix of objects (numerical vector as one object) whose depth is to be calculated; each row contains a `d`-variate point. Should have the same dimension as `data`.
`data`	Matrix of data where each row contains a `d`-variate point, w.r.t. which the depth is to be calculated.

Details

Calculates the convex hull peeling depth (Eddy, 1982; see also Cascos, 2009).

Value

Numerical vector of depths, one for each row in x; or one depth value if x is a numerical vector. Each depth value equals the number of the convex hulls to be peeled from data so that (the corresponding row of) x is not contained in the convex hull of the rest of the data; the depths are normalized by the number of points in data.

References

Eddy, W.F. (1982). Convex hull peeling. In: Caussinus, H., Ettinger, P. and Tomassone, R. (eds), COMPSTAT 1982. Proceedings in computational statistics, Physica-Verlag (Vienna), 42–47.

Cascos, I. (2009). Data depth: multivariate statistics and geometry. In: Kendall, W.S. and Molchanov, I. (eds) New Perspectives in Stochastic Geometry, Clarendon/Oxford University Press (Oxford).

Examples

# Mixture of 3-variate normal distributions
data <- mvrnorm(25, rep(0, 3), diag(3))
x <- rbind(mvrnorm(10, rep(1, 3), diag(3)), data)
depths <- depth.qhpeeling(x, data)
cat("Depths:", depths, "\n")

ddalpha documentation built on Oct. 1, 2024, 1:07 a.m.