swd: Decomposition

Description

This function performs decomposition with multi-sale SBF's.

Usage

swd(sbf)

Arguments

sbf

an object of class `sbf'

Value

An object of class spherical wavelet decomposition(`swd'). This object is a list with the following components.
obsobservations
latlongrid points of observation sites in degree
netlabvector of labels representing sub-networks
etabandwidth parameters for Poisson kernel
methodextrapolation methods, `"ls"' or `"pls"'
approxif TRUE, approximation is used.
grid.sizegrid size (latitude, longitude) of extrapolation site
lambdasmoothing parameter for penalized least squares method
p0starting level for extrapolation. Resolution levels $p0+1, \ldots, L$ is used for extrapolation.
gridlonlongitudes of extrapolation sites in degree
gridlatlatitudes of extrapolation sites in degree
nlevelsthe number of multi-resolution levels
coeffinterpolation coefficients
fieldextrapolation on grid.size
density1density of SBF
latlimrange of latitudes in degree
lonlimrange of longitudes in degree
globalList of successively smoothed data
densitydensity of SW coefficients
detailList of details at different resolution levels
swcoeffSW coefficients
thresh.info`"None"'

Details

This function performs decomposition with multi-sale SBF's.

References

Oh, H-S. and Li, T-H. (2004) Estimation of global temperature fields from scattered observations by a spherical-wavelet-based spatially adaptive method. Journal of the Royal Statistical Society Ser. B, 66, 221--238.

Examples

Run this code

### Observations of year 1967
#data(temperature)
#names(temperature)

# Temperatures on 939 weather stations of year 1967    
#temp67 <- temperature$obs[temperature$year == 1967] 
# Locations of 939 weather stations    
#latlon <- temperature$latlon[temperature$year == 1967, ]

### Network design by BUD
#data(netlab)

### Bandwidth for Poisson kernel
#eta <- c(0.961, 0.923, 0.852, 0.723, 0.506)

### SBF representation of the observations by pls
#out.pls <- sbf(obs=temp67, latlon=latlon, netlab=netlab, eta=eta, 
#    method="pls", grid.size=c(50, 100), lambda=0.89)

### Decomposition
#out.dpls <- swd(out.pls)

Run the code above in your browser using DataLab