CovEst.2010OAS: Oracle Approximating Shrinkage Estimator

In CovTools: Statistical Tools for Covariance Analysis

Description

Authors propose to estimate covariance matrix by iteratively approximating the shrinkage with

\hat{Σ} = ρ \hat{F} + (1-ρ) \hat{S}

where ρ \in (0,1) a control parameter/weight, \hat{S} an empirical covariance matrix, and \hat{F} a target matrix. It is proposed to use a structured estimate \hat{F} = \textrm{Tr} (\hat{S}/p) \cdot I_{p\times p} where I_{p\times p} is an identity matrix of dimension p.

Usage

CovEst.2010OAS(X)

Arguments

X	an (n\times p) matrix where each row is an observation.

Value

a named list containing:

S

a (p\times p) covariance matrix estimate.

rho

an estimate for convex combination weight.

References

\insertRef

chen_shrinkage_2010CovTools

Examples

## CRAN-purpose small computation
# set a seed for reproducibility
set.seed(11)

#  small data with identity covariance
pdim      <- 5
dat.small <- matrix(rnorm(10*pdim), ncol=pdim)

#  run the code
out.small <- CovEst.2010OAS(dat.small)

#  visualize
opar <- par(no.readonly=TRUE)
par(mfrow=c(1,3), pty="s")
image(diag(pdim)[,pdim:1],     main="true cov")
image(cov(dat.small)[,pdim:1], main="sample cov")
image(out.small$S[,pdim:1],    main="estimated cov")
par(opar)

## Not run: 
## want to see how delta is determined according to
#  the number of observations we have.
nsamples = seq(from=5, to=200, by=5)
nnsample = length(nsamples)

#  we will record two values; rho and norm difference
vec.rho   = rep(0, nnsample)
vec.normd = rep(0, nnsample)
for (i in 1:nnsample){
  dat.norun <- matrix(rnorm(nsamples[i]*pdim), ncol=pdim) # sample in R^5
  out.norun <- CovEst.2010OAS(dat.norun)                  # run with default

  vec.rho[i]   = out.norun$rho
  vec.normd[i] = norm(out.norun$S - diag(pdim),"f")       # Frobenius norm
}

# let's visualize the results
opar <- par(no.readonly=TRUE)
par(mfrow=c(1,2))
plot(nsamples, vec.rho,   lwd=2, type="b", col="red", main="estimated rhos")
plot(nsamples, vec.normd, lwd=2, type="b", col="blue",main="Frobenius error")
par(opar)

## End(Not run)

CovTools documentation built on Aug. 14, 2021, 1:08 a.m.