PsiJ: Compute discrete autocorrelation wavelets. In LS2W: Locally Stationary Two-Dimensional Wavelet Process Estimation Scheme

Description

This function computes discrete autocorrelation wavelets.

The inner products of the discrete autocorrelation wavelets are computed by the routine ipndacw.

Usage

 1 2 PsiJ(J, filter.number = 10, family = "DaubLeAsymm", tol = 1e-100, OPLENGTH = 2000, verbose = FALSE)

Arguments

 J Discrete autocorrelation wavelets will be computed for scales -1 up to scale J. This number should be a negative integer. filter.number The index of the wavelet used to compute the discrete autocorrelation wavelets. family The family of wavelet used to compute the discrete autocorrelation wavelets. tol In the brute force computation for Daubechies compactly supported wavelets many inner product computations are performed. This tolerance discounts any results which are smaller than tol which effectively defines how long the inner product/autocorrelation products are. OPLENGTH This integer variable defines some workspace of length OPLENGTH. The code uses this workspace. If the workspace is not long enough then the routine will stop and probably tell you what OPLENGTH should be set to. verbose If this TRUE certain informative statements are printed to the screen when this function is being executed.

Details

This function computes the discrete autocorrelation wavelets. It does not have any direct use for time-scale analysis (e.g. ewspec). However, it is useful to be able to numerically compute the discrete autocorrelation wavelets for arbitrary wavelets and scales as there are still unanswered theoretical questions concerning the wavelets. The method is a brute force – a more elegant solution would probably be based on interpolatory schemes.

Horizontal scale. This routine returns only the values of the discrete autocorrelation wavelets and not their horizontal positions. Each discrete autocorrelation wavelet is compactly supported with the support determined from the compactly supported wavelet that generates it. See the paper by Nason, von Sachs and Kroisandt which defines the horizontal scale (but basically the finer scale discrete autocorrelation wavelets are interpolated versions of the coarser ones. When one goes from scale j to j-1 (negative j remember) an extra point is inserted between all of the old points and the discrete autocorrelation wavelet value is computed there. Thus as J tends to negative infinity the numerical approximation tends towards the continuous autocorrelation wavelet.

This function stores any discrete autocorrelation wavelet sets that it computes. The storage mechanism is not as advanced as that for ipndacw and its subsidiary routines rmget and firstdot but helps a little bit. The Psiname function defines the naming convention for objects returned by this function.

Sometimes it is useful to have the discrete autocorrelation wavelets stored in matrix form. The PsiJmat does this.

Value

A list containing -J components, numbered from 1 to -J. The [[j]]th component contains the discrete autocorrelation wavelet at scale j.

Idris Eckley

References

Eckley, I.A., Nason, G.P. and Treloar, R.L. (2010) Locally stationary wavelet fields with application to the modelling and analysis of image texture. Journal of the Royal Statistical Society (Series C), 59, 595 - 616.

Eckley, I.A. and Nason, G.P. (2011). LS2W: Implementing the Locally Stationary 2D Wavelet Process Approach in R, Journal of Statistical Software, 43(3), 1-23. URL http://www.jstatsoft.org/v43/i03/.