"spatial.neighbor"
This class of objects is constructed using the function
spatial.neighbor
.
Alternatively, the functions
read.neighbor
, or
neighbor.grid
may be used. In general, the user must construct
these objects whenever estimates of spatial correlation and spatial regression
are desired.
An object of class
"spatial.neighbor"
contains all the information
required to determine which spatial units on a region of interest are
neighbors, as well as the strength of their relationship.
The class
"spatial.neighbor"
has associated methods,
print.spatial.neighbor
,
plot.spatial.neighbor
, and
summary.spatial.neighbor
.
Class
"spatial.neighbor"
inherits from class
"data.frame"
.
The
"spatial.neighbor"
object is in essence a data frame with
additional attributes. Each row of the data frame denotes a pair of
neighboring spatial units. The data frame contains the
following columns:
1
to the
total number of regions.
row.id
.
row.id
and
col.id
might not include ALL the spatial units
in the area of interest. This happens when units are isolated, i.e. have
no neighboring regions. In this case,
nregion
must be used to determine the total number of rows and
columns in the contiguity matrix.
TRUE
) or not (
FALSE
).
If
TRUE
, only the weights for the upper (or lower)
triangle of the contiguity matrix need to be specified in the object. Use the
function
spatial.weights
to expand the full symmetric weights
matrix.
An object of class
"spatial.neighbor"
is a sparse matrix representation of a
square matrix (or a number of square matrices).
The function
plot.spatial.neighbor
will show a graphical view
of the spatial.neighbor object
and
summary.spatial.neighbor
will compute summary statistics on the object.
The functions
spatial.multiply
, and
spatial.cg.solve
can be used to form
products of the form
rho[i]*N[i]*x
and
(rho[i]*N[i])^(-1)*x
,
for neighbor weight matrices
N[i]
, vector of constants or parameters,
rho[i]
, and arbitrary vectors
x
, should that be needed to form a
neighbor or contiguity matrix as a weighted linear combination of others.