orderMaxMinFaster: Faster maximin ordering by Euclidean distance

In katzfuss-group/NPvecchia: Large Scale Covariance Estimation

Faster maximin ordering by Euclidean distance

Description

Assumes Euclidean distance and finds maximin ordering. It is equivalent to order_maximin_dist but much faster.

Usage

orderMaxMinFaster(locs, low_mem = FALSE)

Arguments

locs	matrix of n locations in arbitrary dimension (to match input argument of fields::rdist)
low_mem	flag that changes the memory requirements of the function. Defaults to false, which allows the function to calculate the full distance matrix between all locations. If set to true, it will only calculate one column of the distance matrix at a time, which slows down the function but also decreases memory requirements.

Value

Maximin ordering of the points

Examples

n <- 100
d <- 3
locationss <- matrix(runif(n * d), nc = d)
ordering <- orderMaxMinFaster(locationss)

katzfuss-group/NPvecchia documentation built on April 15, 2022, 2:23 a.m.