precmat.GMRFreglat(n,m, par, model = "m1p1", eps = .Spam$eps)
Arguments
n
first dimension of the grid.
m
second dimension of the grid.
par
parameters used to construct the matrix.
model
see details and examples.
eps
A tolerance parameter: elements of x such that abs(x)
<= eps set to zero. Defaults to eps = .Spam$eps
Details
The function should be illustrative on how to create precision
matrices for gridded GMRF. Hence, no testing (positive definiteness is
done).
The model specification "m" determines the complexity and
"p" the number of parameters.
Please see the examples on the meaning of the different models.
Value
A spam matrix of dimension prod(dims)xprod(dims).
Author(s)
Reinhard Furrer
See Also
precmat, toeplitz.spam, kronecker.spam
Examples
as.matrix(precmat.GMRFreglat(4, 3, c(.4), 'm1p1'))
as.matrix(precmat.GMRFreglat(4, 3, c(.4,.3), 'm1p2'))
as.matrix(precmat.GMRFreglat(4, 3, c(.4,.3,.2), 'm2p3'))
as.matrix(precmat.GMRFreglat(4, 3, c(.4,.3,.2,.1),'m2p4'))
# up to the diagonal, the following are equivalent:
cleanup( precmat.IGMRFreglat(3,4) -
precmat.GMRFreglat(3,4,1, 'm1p1'))