covI {locits} | R Documentation |
Compute the covariance between two wavelet periodogram ordinates at the same scale, but different time locations.
Description
Computes cov(I_{\ell, m}, I_{\ell, n})
using the formula
given in Nason (2012) in Theorem 1. Note: one usually should
use the covIwrap
function for efficiency.
Usage
covI(II, m, n, ll, ThePsiJ)
Arguments
II |
Actually the *spectral* estimate S, not the periodogram values. This is for an assumed stationary series, so this is just a vector of length J, one for each scale of S. |
m |
Time location m |
n |
Time location n |
ll |
Scale of the raw wavelet periodogram |
ThePsiJ |
Autocorrelation wavelet corresponding to the wavelet that computed the raw peridogram (also assumed to underlie the time series |
Value
The covariance is returned.
Author(s)
Guy Nason.
References
Nason, G.P. (2013) A test for second-order stationarity and approximate confidence intervals for localized autocovariances for locally stationary time series. J. R. Statist. Soc. B, 75, 879-904. doi:10.1111/rssb.12015
See Also
Examples
P1 <- PsiJ(-5, filter.number=1, family="DaubExPhase")
#
# Compute the covariance
#
covI(II=c(1/2, 1/4, 1/8, 1/16, 1/32), m=1, n=3, ll=5, ThePsiJ=P1)
#
# [1] 0.8430809