tguh.decomp {breakfast}R Documentation

The Tail-Greedy Unbalanced Haar decomposition of a vector

Description

This function performs the Tail-Greedy Unbalanced Haar decomposition of the input vector.

Usage

tguh.decomp(x, p = 0.01)

Arguments

x

A vector you wish to decompose.

p

Specifies the number of region pairs merged in each pass through the data, as the proportion of all remaining region pairs. The default is 0.01.

Details

The Tail-Greedy Unbalanced Haar decomposition algorithm is described in "Tail-greedy bottom-up data decompositions and fast multiple change-point detection", P. Fryzlewicz (2017), preprint.

Value

A list with the following components:

n

The length of x.

decomp.hist

The decomposition history: the complete record of the n-1 steps taken to decompose x. This is an array of dimensions 4 by 2 by n-1. Each of the n-1 matrices of dimensions 4 by 2 contains the following: first row - the indices of the regions merged, in increasing order (note: the indexing changes through the transform); second row - the values of the Unbalanced Haar filter coefficients used to produce the corresponding detail coefficient; third row - the (detail coefficient, smooth coefficient) of the decomposition; fourth row - the lengths of (left wing, right wing) of the corresponding Unbalanced Haar wavelet.

tguh.coeffs

The coefficients of the Tail-Greedy Unbalanced Haar transform of x.

Author(s)

Piotr Fryzlewicz, p.fryzlewicz@lse.ac.uk

See Also

tguh.cpt, tguh.denoise, tguh.reconstr

Examples

rnoise <- rnorm(10)
tguh.decomp(rnoise)

[Package breakfast version 1.0.0 Index]