Tools for analysing multivariate time series with wavelets. This includes: simulation of a multivariate locally stationary wavelet (mvLSW) process from a multivariate evolutionary wavelet spectrum (mvEWS); estimation of the mvEWS, local coherence and local partial coherence. See Park, Eckley and Ombao (2014) for details.

Documentation

Manual: mvLSW.pdf
Vignette: None available.

Maintainer: Simon Taylor <s.taylor2 at lancaster.ac.uk>

Author(s): Simon Taylor*, Tim Park*, Idris Eckley*, Rebecca Killick*

Install package and any missing dependencies by running this line in your R console:

install.packages("mvLSW")

Depends R (>= 3.2), fields, wavethresh, xts, zoo
Imports
Suggests
Enhances
Linking to
Reverse
depends
Reverse
imports
Reverse
suggests
Reverse
enhances
Reverse
linking to

Package mvLSW
Materials
URL
Task Views
Version 1.2
Published 2017-08-06
License GPL (>= 3)
BugReports
SystemRequirements
NeedsCompilation yes
Citation
CRAN checks mvLSW check results
Package source mvLSW_1.2.tar.gz