Fix model parameters

Share:

Description

This function fixes theta (model parameter vector) after optimization. That is, if a value was optimized on a logarithmic scale: theta=10^theta. The result depends on dimensionality of the problem, the nugget value and the choice of correlation function.

Usage

1
daceFixTheta(m, bestTheta, nugget, corr)

Arguments

m

dimension (length) of each point

bestTheta

the theta value found during MLE

nugget

Value for nugget. Default is -1, which means the nugget was be optimized during MLE. In this case, it is part of bestTheta. Otherwise, nugget is appended to the theta vector. This is only relevant for correlation functions that use a nugget (corrnoisygauss,corrnoisykriging )

corr

The choice of correlation function (which defines the length and values of theta and bounds): corrnoisykriging (default), corrkriging, corrnoisygauss, corrgauss, correxp, correxpg, corrlin, corrcubic,corrspherical,corrspline. Can also be user supplied (if in the right form).

Value

returns a list: thetaConv the fixed theta vector (all model parameters) theta vector of activity parameters theta p vector of exponents p(NULL if not used in correlation function) lambda nugget value lambda (NULL if not used in correlation function)

See Also

daceBuilder

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.