bivquant (version 0.1)

geomqu: On Affine Equivariant Multivariate Quantiles

Description

This function fits empirical bivariate quantiles as proposed by Chakraborty, B. (2001) <https://doi.org/10.1023/A:1012478908041>.

Usage

geomqu(data, u, p, alpha, keep_optim = FALSE, keep_data = TRUE)

Arguments

data

must be a (n,d) matrix of observations.

u

a (m,d) matrix with m directions.

p

defines the p-norm, must be in \([1,\infty)\).

alpha

must be missing or a vector of length d+1 with distict values from 1 to n.

keep_optim

boolean keep results from optimization in result object, default is FALSE.

keep_data

boolean keep data parameter in result object, default is TRUE.

Value

an object of class geomqu with methods plot.geomqu and plot.geomqu.

References

Chakraborty, B. (2001). On affine equivariant multivariate quantiles. Annals of the Institute of Statistical Mathematics, 53, 380--. <https://doi.org/10.1023/A:1012478908041>.