CVTuningCov (version 1.0)

hard.thresholding: Hard-thresholding Operator on A Covariance Matrix

Description

Apply hard-thresholding operator on an input covariance matrix with a tuning parameter.

Usage

hard.thresholding(Sigma, c = 0.5)

Arguments

Sigma

a covariance matrix with dimension p*p.

c

the default value is 0.5.

Value

a p*p covariance matrix after hard-thresholding operation.

References

Bickel, P and Levina, E, Covariance regularization by thresholding, Annals of Statistics, 36, 2577-2604 (2008).

Examples

p <- 5;
Sigma <- AR1(p, rho=0.6);
hard.Sigma<-hard.thresholding(Sigma,c=0.5);
hard.Sigma;