makegaussiankernel: function to make a gaussian kernel

In ethanplunkett/gridkernel: Package to apply kernels to raster data with NA and edge correction

function to make a gaussian kernel

Description

this function makes a kernel where the value in each cell is a gaussian function of the distance between that cell and the center of the kernel.

Usage

makegaussiankernel(sd, max.r = 3 * sd, cellsize = 1, include.center.cell = TRUE)

Arguments

sd	the standard deviation in map units of the gaussian function
max.r	the maximum radius of the kernel in map units. This defaults to three standard deviations. A smaller radius speeds calculation but loses information at the tail.
cellsize	the size of each cell in the grid in map units
include.center.cell	if FALSE the value of the center cell will be forced to zero. This was helpful in my original use of the function but probably not to you.

Value

a square, numeric matrix which sums to 1 and in which the value of each cell is a gaussian function of the distance from the cell to the center of the matrix.

Author(s)

Ethan Plunkett

See Also

calckernel, kernelsmooth

Examples

k <- makegaussiankernel(sd=2)
m <- matrix(1:60, 6, 10)
sm <- kernelsmooth(m, k)

ethanplunkett/gridkernel documentation built on March 2, 2024, 9:06 p.m.