compute_distance_matrix: Compute distance matrix

In GWlasso: Geographically Weighted Lasso

Compute distance matrix

Description

compute_distance_matrix() is a small helper function to help you compute a distance matrix. For the geographically method to work, is is important that distances between points are not zero. This function allows to add a small random noise to avoid zero distances.

Usage

compute_distance_matrix(data, method = "euclidean", add.noise = FALSE)

Arguments

data	A dataframe or matrix containing at least two numerical columns.
method	method to compute the distance matrix. Ultimately passed to stats::dist(). Can be euclidean, maximum, manhattan, canberra, binary or minkowski.
add.noise	TRUE/FALSE set to TRUE to add a small noise to the distance matrix. Noise U is generated as U \sim (1\times 10^{-6}, 5\times 10^{-6}). Noise is added only for pairs for which distance is zero.

Value

a distance matrix, usable in gwl_bw_estimation()

Examples

coords <- data.frame("Lat" = rnorm(200), "Long" = rnorm(200))
distance_matrix <- compute_distance_matrix(coords)

GWlasso documentation built on April 3, 2025, 7:08 p.m.