learnParameters: Projected Sequential Gaussian Process

In psgp: Projected Spatial Gaussian Process Methods

Projected Sequential Gaussian Process

Description

learnParameters performs maximum likelihood parameter estimation in the PSGP framework.

Usage

  learnParameters(object)

Arguments

object

a list object of intamap type. Most arguments necessary for interpolation are passed through this object. See intamap-package for further description of the necessary content of this variable.

Details

The Projected Spatial Gaussian Process (PSGP) framework provides an approximation to the full Gaussian process in which the observations are projected sequentially onto an optimal subset of 'active' observations. Spatial interpolation is done using a mixture of covariance kernels (exponential and Matern 5/2).

The function learnParameters is an internal function for estimating the parameters of the covariance function given the data, using a maximum likelihood approach. A valid intamap object must be passed in.

PSGP is able to also take the measurement characteristics (i.e. errors) into account using possibly many error models. For each error model, assumed Gaussian, the error variance can be specified. The vector object$observations$oevar contains all variances for the error models (one value per error model). Which error model is used for a given observation is determined by the object$observations$oeid vector of indices, which specifies the index of the model to be used for each observation.

Warning

It is advised to use the intamap wrapper estimateParameters rather than calling this method directly.

Author(s)

Ben Ingram, Remi Barillec

References

Csato and Opper, 2002; Ingram et al., 2008

See Also

makePrediction, learnParameters, estimateParameters, createIntamapObject

Examples

  # see example in estimateParameters

psgp documentation built on Nov. 27, 2023, 5:09 p.m.