km-class {DiceKriging}R Documentation

Kriging models class

Description

S4 class for kriging models.

Objects from the Class

To create a km object, use km. See also this function for more details.

Slots

d:
Object of class "integer". The spatial dimension.
n:
Object of class "integer". The number of observations.
X:
Object of class "matrix". The design of experiments.
y:
Object of class "matrix". The vector of response values at design points.
p:
Object of class "integer". The number of basis functions of the linear trend.
F:
Object of class "matrix". The experimental matrix corresponding to the evaluation of the linear trend basis functions at the design of experiments.
trend.formula:
Object of class "formula". A formula specifying the trend as a linear model (no response needed).
trend.coef:
Object of class "numeric". Trend coefficients.
covariance:
Object of class "covTensorProduct". See covTensorProduct-class.
noise.flag:
Object of class "logical". Are the observations noisy?
noise.var:
Object of class "numeric". If the observations are noisy, the vector of noise variances.
known.param:
Object of class "character". Internal use. One of: "None", "All" or "Trend".
case:
Object of class "character". Internal use. One of: "NoNugget", "1Nugget", "Nuggets".
param.estim:
Object of class "logical". TRUE if at least one parameter is estimated, FALSE otherwise.
method:
Object of class "character". "MLE" or "PMLE" depending on penalty.
penalty:
Object of class "list". For penalized ML estimation.
optim.method:
Object of class "character". To be chosen between "BFGS" and "gen".
lower:
Object of class "numeric". Lower bounds for covariance parameters estimation.
upper:
Object of class "numeric". Upper bounds for covariance parameters estimation.
control:
Object of class "list". Additional control parameters for covariance parameters estimation.
gr:
Object of class "logical". Do you want analytical gradient to be used ?
call:
Object of class "language". User call reminder.
parinit:
Object of class "numeric". Initial values for covariance parameters estimation.
logLik:
Object of class "numeric". Value of the concentrated log-Likelihood at its optimum.
T:
Object of class "matrix". Triangular matrix delivered by the Choleski decomposition of the covariance matrix.
z:
Object of class "numeric". Auxiliary variable: see computeAuxVariables.
M:
Object of class "matrix". Auxiliary variable: see computeAuxVariables.

Methods

plot
signature(x = "km"): see plot.km.
predict
signature(object = "km"): see predict.km.
show
signature(object = "km"): see show.km.
simulate
signature(object = "km"): see simulate.km.

Author(s)

O. Roustant, D. Ginsbourger

See Also

km for more details about slots and to create a km object, covStruct.create to construct a covariance structure, and covTensorProduct-class for the S4 covariance class defined in this package.


[Package DiceKriging version 1.0 Index]