soft.thresholding {CVTuningCov} | R Documentation |
Apply soft-thresholding operator on an input covariance matrix with a tuning parameter.
soft.thresholding(Sigma, c = 0.5)
Sigma |
a covariance matrix with dimension |
c |
the default value is 0.5. |
a p*p
covariance matrix after soft-thresholding operation.
Binhuan Wang
Rothman, A, Levina, E and Zhu, J, A new approach to Cholesky-based covariance regularization in high dimensions, Biometrika, 97, 539-550 (2010).
p <- 5; Sigma <- AR1(p, rho=0.6); soft.Sigma<-soft.thresholding(Sigma,c=0.5); soft.Sigma;