network.design: Network Design

In SpherWave: Spherical Wavelets and SW-based Spatially Adaptive Methods

Description

produces multi-resolution network.

Usage

network.design(latlon, method = "Oh", type = "reduce", nlevel, x) 

Arguments

latlon	grid points of observation sites in degree
method	network design method, "cover", "ModifyGottlemann", "Gottlemann" or "Oh"
type	specifies grid type, "regular" or "reduced"
nlevel	the number of observations in each resolution when using the method ‘"cover"’
x	radius in degree

Details

This function partitions the grid points of observations into networks. Each network corresponds resolution level and level 1 is the most detailed level. Possible methods are

‘"cover"’ for utilizing ‘"cover.design"’ in the package ‘"field"’

‘"ModifyGottlemann"’ for modified G\"ottlemann's method

‘"Gottlemann"’ for G\"ottlemann's method

‘"Oh"’ for Oh's method.

For ‘"ModifyGottlemann"’, ‘"Gottlemann"’ and ‘"Oh"’, two types of design, ‘"regular"’ and ‘"reduced"’ are provided.

Value

netlab

vector of network labels indicating level of multi-resolution.

References

Oh, H-S. (1999) Spherical wavelets and their statistical analysis with applications to meteorological data. Ph.D. Thesis, Department of Statistics, Texas A\&M University, College Station.

Examples

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

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

#netlab <- network.design(latlon=latlon, method="cover",
#    nlevel=c(507, 244, 117, 49, 22))
#netlab <- network.design(latlon=latlon, method="ModifyGottlemann", 
#    type="regular", x=3)
#netlab <- network.design(latlon=latlon, method="Gottlemann", 
#    type="regular", x=2)
#netlab <- network.design(latlon=latlon, method="Oh", 
#    type="reduce", x=5)

SpherWave documentation built on April 14, 2017, 1:28 p.m.