elem_autocov: Element-wise tail-robust autocovariance matrix estimator

In HaotianXu/rcov: Robust Autocovariance Matrix Estimation

Element-wise tail-robust autocovariance matrix estimator

Description

Compute the element-wise tail-robust matrix estimator.

Usage

elem_autocov(X, l, est.type = "truncate", tau.mat, block.size, ...)

Arguments

X	A numeric p x n matrix of observations.
l	A non-negative integer corresponding to the lag of the autocovariance.
est.type	A character string, such as "truncate", "huber" or "MoM", indicating the type of element-wise tail-robust estimator.
tau.mat	A numeric p x p matrix of robustification parameters (larger than 0), when est.type is "truncate" or "huber".
block.size	A strictly positive integer scalar corresponding to the block size, when est.type is "MoM".
...	Additional arguments.

Value

A numeric scalar corresponding to the median-of-means estimator.

Author(s)

Haotian Xu

Examples

set.seed(123)
X = matrix(rt(50*100, 3), nrow = 50, ncol = 100)
gamma1_trunc = elem_autocov(X, l = 1, est.type = 'truncate', tau.mat = matrix(5, nrow = 50, ncol = 50))
gamma1_huber = elem_autocov(X, l = 1, est.type = 'huber', tau.mat = matrix(5, nrow = 50, ncol = 50))
gamma1_MoM = elem_autocov(X, l = 1, est.type = 'MoM', block.size = 10)
dim(gamma1_trunc)
dim(gamma1_huber)
dim(gamma1_MoM)

HaotianXu/rcov documentation built on May 14, 2023, 5:04 a.m.