awg_knots: Augment a set of knots with a grid

In jelsema/RRSM: Methods for reduced rank spatial models

Description

Takes a set of knot locations and augments it with a grid, so that 'empty' areas are filled.

Usage

awg_knots(init_knots, xdim = c(0, 100), ydim = c(0, 100), kclose = 2)

Arguments

init_knots	the initial set of knots, e.g., selected using TKS.
xdim	the x-coordinate limits of the spatial domain
ydim	the y-coordinate limits of the spatial domain
kclose	the number of other knots to consider as the ‘neighborhood’ of a knot. See details.

Details

awg_knots computes the average distance to the next kclose knots, and creates a grid of knots such that the minimum distance is approximately equal to this value. More precisely, let U be the set of initial knots, and U[j,] be a particular knot. Then if kclose = 2 , the pairwise distances between U[j,] and U[-j,] are computed, and the smallest two are averaged. Suppose a vector d contains all these mean distances d[j], for j = 1, ..., nrow(init_knots). Then mean(d) is the average distance to the next kclose knots.

Value

A matrix containing a set of locations.

See Also

TKS

jelsema/RRSM documentation built on May 19, 2019, 4:02 a.m.