kriging {kriging}

Ordinary Kriging

Description

Simple and highly optimized ordinary kriging algorithm to plot geographical data

Usage

kriging(x, y, response, model = "spherical", lags = 10, pixels = 100, polygons = NULL)

Arguments

x

vector of x-axis spatial points.

y

vector of y-axis spatial points.

response

vector of observed values.

model

specification of the variogram model. Choices are "spherical", "exponential" or "gaussian". Defaults to "spherical".

lags

number of lags. Defaults to 10.

pixels

maximum number of points along either axis. Defaults to 100.

polygons

list of polygons used to grid predicted values on to. The default value of NULL automatically generates an evenly spaced out rectangular grid of points spanning the range of the data.

Details

The kriging algorithm assumes a minimum number of observations in order to fit the variogram model.

Value

An object of class kriging that inherits from list and is composed of:

model

character; variogram model.

nugget

numeric; value of nugget parameter.

range

numeric; value of range parameter.

sill

numeric; value of sill parameter.

map

data.frame; contains the predicted values along with the coordinate covariates.

semivariogram

data.frame; contains the distance and semivariance values.

Author(s)

Omar E. Olmedo

See Also

image.kriging, plot.kriging.

Examples

# Krige random data for a specified area using a list of polygons
library(maps)
usa <- map("usa", "main", plot = FALSE)
p <- list(data.frame(usa$x, usa$y))

# Create some random data
x <- runif(50, min(p[[1]][,1]), max(p[[1]][,1]))
y <- runif(50, min(p[[1]][,2]), max(p[[1]][,2]))
z <- rnorm(50)

# Krige and create the map
kriged <- kriging(x, y, z, polygons=p, pixels=300)
image(kriged, xlim = extendrange(x), ylim = extendrange(y))