kmPredictWrapper {fanovaGraph}

Wrapper for the Kriging model prediction

Description

Wrapper for the Kriging model prediction function predict.km from package DiceKriging to simplify the use of Kriging prediction functions as arguments for functions like estimateGraph or fast99.

Usage

kmPredictWrapper(newdata, km.object)

Arguments

newdata

a vector, matrix or data frame containing the points where to perform predictions

km.object

an object of class km

Value

kriging mean computed at newdata

Author(s)

J. Fruth, O. Roustant

See Also

estimateGraph

Examples

### graph estimation of a kriging prediction of the ishigami function
set.seed(1)
x <- matrix(runif(150,-pi,pi),100,3)
y <- ishigami.fun(x)
KM <- km(~1, design = data.frame(x), response = y)

g <- estimateGraph(f.mat = kmPredictWrapper, d = 3, n.tot = 10000, q.arg = 
  list(min = -pi, max = pi), method = "LiuOwen", km.object = KM) 
print(g$tii)