corSpatial {nlme} | R Documentation |
Spatial Correlation Structure
Description
This function is a constructor for the corSpatial
class,
representing a spatial correlation structure. This class is "virtual",
having four "real" classes, corresponding to specific spatial
correlation structures, associated with it: corExp
,
corGaus
, corLin
, corRatio
, and
corSpher
. The returned object will inherit from one of these
"real" classes, determined by the type
argument, and from the
"virtual" corSpatial
class. Objects created using this
constructor must later be initialized using the appropriate
Initialize
method.
Usage
corSpatial(value, form, nugget, type, metric, fixed)
Arguments
value |
an optional vector with the parameter values in
constrained form. If |
form |
a one sided formula of the form |
nugget |
an optional logical value indicating whether a nugget
effect is present. Defaults to |
type |
an optional character string specifying the desired type of
correlation structure. Available types include |
metric |
an optional character string specifying the distance
metric to be used. The currently available options are
|
fixed |
an optional logical value indicating whether the
coefficients should be allowed to vary in the optimization, or kept
fixed at their initial value. Defaults to |
Value
an object of class determined by the type
argument and also
inheriting from class corSpatial
, representing a spatial
correlation structure.
Author(s)
José Pinheiro and Douglas Bates bates@stat.wisc.edu
References
Cressie, N.A.C. (1993), "Statistics for Spatial Data", J. Wiley & Sons.
Venables, W.N. and Ripley, B.D. (2002) "Modern Applied Statistics with S", 4th Edition, Springer-Verlag.
Littel, Milliken, Stroup, and Wolfinger (1996) "SAS Systems for Mixed Models", SAS Institute.
See Also
corExp
,
corGaus
,
corLin
,
corRatio
,
corSpher
,
Initialize.corStruct
,
summary.corStruct
,
dist
Examples
sp1 <- corSpatial(form = ~ x + y + z, type = "g", metric = "man")