network.design {SpherWave} | R Documentation |
produces multi-resolution network.
network.design(latlon, method = "Oh", type = "reduce", nlevel, x)
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 |
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"{o}ttlemann's method
`"Gottlemann"' for G"{o}ttlemann's method
`"Oh"' for Oh's method.
For `"ModifyGottlemann"', `"Gottlemann"' and `"Oh"', two types of design, `"regular"' and `"reduced"' are provided.
netlab |
vector of network labels indicating level of multi-resolution. |
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.
### 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)