accessD.wst: Get mother wavelet coefficients from a packet ordered...

In wavethresh: Wavelets Statistics and Transforms

Get mother wavelet coefficients from a packet ordered non-decimated wavelet object (wst).

Description

The mother wavelet coefficients from a packet ordered non-decimated wavelet object, wst, are stored in a matrix. This function extracts all the coefficients corresponding to a particular resolution level.

Usage

## S3 method for class 'wst'
accessD(wst, level, aspect = "Identity", ...)

Arguments

wst	Packet ordered non-decimated wavelet object from which you wish to extract the mother wavelet coefficients.
level	The level that you wish to extract. This can range from zero (the coarsest coefficients) to nlevelsWT(wstobj) which returns the original data.
aspect	Function to apply to coefficient before return. Supplied as a character argument which gets converted to a function. For example, "Mod" which returns the absolute value of complex-valued coefficients.
...	Other arguments

Details

The wst function performs a packet-ordered non-decimated wavelet transform. This function extracts all the mother wavelet coefficients at a particular resolution level specified by level.

Note that coefficients returned by this function are in packet order. They can be used as is but for many applications it might be more useful to deal with the coefficients in packets: see the function getpacket.wst for further details.

Note that all the coefficients here are those of mother wavelets. The non-decimated transform efficiently computes all possible shifts of the discrete wavelet transform computed by wd.

Value

A vector of the extracted coefficients.

Author(s)

G P Nason

References

Nason, G.P. and Silverman, B.W. The stationary wavelet transform and some statistical applications.

See Also

wst, wst.object, accessD, getpacket.wst

Examples

#
# Get the 4th level of mother wavelet coefficients from a decomposition
#
dat <- rnorm(128)
accessD(wst(dat), level=4)

wavethresh documentation built on Sept. 11, 2024, 9:33 p.m.