parallel.fcn: Function for paralleled estimation of a Gaussian process...

In ashtonwiens/gp.nonrigid.registration: An R package for rigid and nonrigid registration of spatial surfaces

Description

Using window i of data, subsamples the data and then performs estimation of a penalized Gaussian process with Matern covariance function for two independent sets of data, where a rigid transformation is embedded, applied to the second, moving data set. Each data set is coordinate-wise de-meaned by the means of the first data set's coordinates.

Usage

parallel.fcn(i, gpnl.obj)

Arguments

i	an integer indicating which window/subset of data to apply the Gaussian process rigid registration model to
gpnl.obj	a gpnl.obj containing data.subsets and est.pars components

Value

a list containing the optimization results, the timing results, and the means of each coordinate used in de-meaning before optimization is performed

See Also

gp.rigid.registrataion gp.nonrigid.registration

ashtonwiens/gp.nonrigid.registration documentation built on April 11, 2020, 12:28 a.m.