negLogLik: Negative Log Likelihood Functions

In OakleyJ/MUCM: Gaussian process emulator methods based on the MUCM toolkit

Description

These functions are minimised to find the optimal value of theta (or phi and sigma in LMC case). negLogLikNugget is used in the univariate case for uncertain input parameters. negLogLikLMC is used when applying the LMC Emulator for Multivariate models

Usage

negLogLik(theta, nugget = NULL, inputs, H, outputs, cor.function, ...)

negLogLikNugget(theta, nugget, inputs, H, outputs, cor.function, ...)

negLogLikLMC(phi, sigma, inputs, H, outputs, cor.function, ...)

Arguments

theta	Initial values for the parameters to be optimized over. (includes values for all the parameters that need to be estimated.) For negLogLik, theta represents just phi, for negLogLikNugget, theta represents just phi and sigma combined.
nugget	For noisy data, a vector giving the observation variance for each training data point.
inputs	A data frame, matrix or vector containing the input values of the training data.
H	A matrix of prior mean regressors for the training data.
outputs	A data frame, matrix or vector containing the output values of the training data. In negLogLikLMCOptim, the outputs should be stacked (either a vector or a matrix with 1 column).
cor.function	Specifies a correlation function used as part of the prior information for the emulator
...	additional arguments to be passed on to correlation functions (see corGaussian)
phi	Phi is the roughness parameter for each input and output
sigma	Sigma is matrix of the covariance between different outputs.

Value

The function returns the negetive log-likelihood of theta

Author(s)

Sajni Malde, Jeremy Oakley

See Also

corGaussian

OakleyJ/MUCM documentation built on May 7, 2019, 9:01 p.m.