gaussiansmooth: function to apply a gaussian smoothing to a grid.

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

function to apply a gaussian smoothing to a grid.

Description

This function returns a grid of the same dimension as the input in which each cell has a gaussian weighted mean of cells around it in the input grid. To speed up processing with large bandwidths (SD) it may upscale the data prior to applying a Gaussian kernel. The amount of upscaling is determined by the kernel.dim argument which sets the target dimension (in cells) of the kernel at the courser resolution. The function can use a weights table to run directly on a landcover grid substituting the cover values with weights as defined in the weights table and with the no.match and na.value arguments. Edge correction is used; NA cells and cells outside of the region are not included in the calculation.

Usage

gaussiansmooth(x, sd, max.r, kernel.dim, ...)
## S3 method for class 'grid'
gaussiansmooth(x, sd, max.r=3*sd, kernel.dim=21, max.p.na=1,
        sample.points, weights, no.match=0, use.old = FALSE,
        na.value = NA, ...)

Arguments

x	x should be a grid object in memory.
sd	The standard deviation of the kernel used to do the smoothing (in map units).
max.r	The maximum radius that weights are calculated for (in map units). It defaults to three times sd.
kernel.dim	The dimension of the kernel in the upscaled grid. It must be an odd number. This controls the amount of upscaling, which is adjusted such that a max.r radius kernel has this cell dimension. Smaller values result in more upscaling and faster processing at the expense of precision. Use Inf for no upscaling. The default value of this parameter is 21.
max.p.na	If the proportion of NA values (from the original resolution) in an upscaled cell exceeds max.p.na than it will receive a NA value during upscaling; otherwise the upscaled cell is given the mean value of all non-NA cells that it covers.
sample.points	(optional) a dataframe or matrix with columns "x" and "y" containing the map unit coordinates of points. If specified the function will return both the samples and a grid (in a list).
weights	(optional) a dataframe or matrix with columns "value" and "weight". If weights is provided than all the cells in the grid will be replaced with their associated weight; any cell whose value is not in the value column will be replaced with the values specified by no.match or na.value. If weights does not contain columns named "value" and "weight" and contains only two columns it will be assumed that the first is "value" and the second is "weight".
no.match	If weights are supplied this determines the value to use when cells in x are not NA and do not match any of the values in weights.
na.value	If weights are supplied this determines the value to use when cells in x are NA. Note if the weights argument contains a value for NA as well that will take precedence.
use.old	if FALSE then the newer Rcpp based implementation is used. Otherwise the old, pure R, version is used. This argument is here to facilitate testing.
...	arguments to be passed to other methods.

Value

If sample.points are not provided then the function will return a grid. If sample.points are provided the grid method will return a two item list: sample containing a vector of sampled values, and grid.

Note

It is left up to the user to make sure that the weights table is complete. Any values that occur in the grid but are not in the "value" column of weights will be replaced with no.match prior to smoothing unless they are NA in which case they will be replaced with na.value.

Author(s)

Ethan Plunkett

See Also

kernelsmooth, makegaussiankernel, upscale, and downscale are all called by this function. kernelsmooth does the same thing but allows for user specified kernels. calckernel allows calculating a kernel weighted average for a single cell. kernsource is similar but works by summing kernels centered on each cell, it may be faster if the non-zero values in the grid are sparse and the SD is small. swap is used internally to substitute values when weights are used.

Examples

# Matrix method
m <- matrix(1:60, 6, 10)
k <- matrix(1, 3, 3)
k[2,2] <- 3
k <- k/sum(k)
m2 <- kernelsmooth(m, k)

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