spde.matrices {RGeostats}

R Documentation

Build and return the Precision matrix

Description

Build and return the Precision matrix

Usage

spde.matrices(dbin=NA, dbout=NA, model, triswitch="nqQ", nostat=NA, gext=NA,
              mesh.dbin=TRUE, mesh.dbout=TRUE, mesh=NA, 
              flag.Q=FALSE, flag.sparse=TRUE, flag.advection=FALSE, verbose = FALSE)

Arguments

dbin	The db-class structure containing the input data.
dbout	The db-class structure which contains the results.
model	The model-class describing the spatial characteristics.
triswitch	Command line for the internal triangulation step. For more information see meshing.
nostat	List of non-stationary parameters. For details see model.param.define.
gext	The 'dbout' may be dilated by gext. This argument designates an array, with its dimension equal to the dimension of the space and which contains the extension in each direction. If not defined, the 'dbout' is not dilated and the simulated results may suffer some edge effect problems.
mesh.dbin	When TRUE, the location corresponding to the Input Data are systematically included in the meshing
mesh.dbout	When TRUE, the location corresponding to the Output Targets are systematically included in the meshing
mesh	A mesh-class object.
flag.Q	When TRUE, the Q sparse matrix is output (in triplet format).
flag.sparse	When TRUE (if the package 'Matrix' is loaded), the resulting matrices are returned as a sparse matrix.
flag.advection	When TRUE, the matrices are established in presence of advection (specified as a non-stationary arameter). Then the matrix which is returned is G instead of Q.
verbose	Verbose flag.

Value

Several vectors or matrices used in the SPDE framework are provided such as:

• Q The precision matrix

• G The G array

• S The S array: C^(-1/2) * G * C^(-1/2)

• TildeC The vector C vector (used for anisotropy and range)

• Lambda The L vector (used for sill): L = (Lb * C)^(1/2)

• Vars The diagonal of the inverse Q matrix

• blin The array of coefficients used to construct Q

• indCo The list of coarse(1) or fine(0) samples

• L The L array (connectivity sparse matrix)

• order Meshing order: 1 (Input,Output,Steiner); 2 (Output,Input,Steiner)

• inddat which gives the indices of the data samples in Q

• nbnodup which gives the number of samples of Dbin which do not coincide with samples of Dbout.

Note that:

• Q is produced if 'flag.Q' is TRUE

• G, S, TildeC and Lambda are produced when calculated

• blin are produced when calculated

The matrices 'Q', 'G' and 'S'are sparse. They are returned in the 'triplet' form if 'flag.sparse' is FALSE. Each non-zero element of the sparse matrix is specified by its row, column and the item value. If 'flag.sparse'=TRUE, the matrics are returned as sparse matrices (of the package 'Matrix').

---