st_kernel_weights: Calculate Kernel Weights

In sfdep: Spatial Dependence for Simple Features

Calculate Kernel Weights

Description

Create a weights list using a kernel function.

Usage

st_kernel_weights(
  nb,
  geometry,
  kernel = "uniform",
  threshold = critical_threshold(geometry),
  adaptive = FALSE,
  self_kernel = FALSE
)

Arguments

nb	an object of class nb e.g. created by st_contiguity() or st_knn().
geometry	the geometry an sf object.
kernel	One of "uniform", "gaussian", "triangular", "epanechnikov", or "quartic". See kernels for more.
threshold	a scaling threshold to be used in calculating
adaptive	default FALSE. If TRUE uses the maximum neighbor distance for each region as the threshold. Suppresses the threshold argument.
self_kernel	default FALSE. If TRUE applies the kernel function to the observed region.

Details

By default st_kernel_weight() utilizes a critical threshold of the maximum neighbor distance using critical_threshold(). If desired, the critical threshold can be specified manually. The threshold will be passed to the underlying kernel.

Value

a list where each element is a numeric vector.

See Also

Other weights: st_inverse_distance(), st_nb_dists(), st_weights()

Examples

geometry <- sf::st_geometry(guerry)
nb <- st_contiguity(geometry)
nb <- include_self(nb)
res <- st_kernel_weights(nb, geometry)
head(res, 3)

sfdep documentation built on Jan. 11, 2023, 9:08 a.m.