covMat1Mat2 {DiceKriging}

Cross covariance matrix

Description

Computes the cross covariance matrix between two sets of locations for a spatial random process with a given covariance structure. Typically the two sets are a learning set and a test set.

Usage

covMat1Mat2(object, X1, X2, nugget.flag=FALSE)

Arguments

object

an object specifying the covariance structure.

X1

a matrix whose rows represent the locations of a first set (for instance a set of learning points).

X2

a matrix whose rows represent the locations of a second set (for instance a set of test points).

nugget.flag

an optional boolean. If TRUE, the covariance between 2 equal locations takes into account the nugget effect (if any). Locations are considered equal if their euclidian distance is inferior to 1e-15. Default is FALSE.

Value

a matrix of size (nb of rows of X1 * nb of rows of X2) whose element (i1,i2) is equal to the covariance between the locations specified by row i1 of X1 and row i2 of X2.

Author(s)

Olivier Roustant, David Ginsbourger, Ecole des Mines de St-Etienne.

See Also

covMatrix