geostatsp (version 0.7.0)

krige: Spatial prediction, or Kriging

Description

Perform spatial prediction, producing a raster of predictions and conditional standard deviations.

Usage

krige(data, trend, coordinates = data, param, locations, 
    covariates = NULL, expPred = FALSE, nugget.in.prediction = TRUE)

Arguments

data

A SpatialPointsDataFrame containing the data to be interpolated, or a vector of observed data

trend

Either a model formula, or a data frame of linear covariates.

coordinates

A SpatialPoints object giving the locations of the observed data. If data is a SpatialPointsDataFrame coordinates is not necessary.

covariates

The spatial covariates used in prediction, either a raster stack or list of rasters.

param

A vector of named model parameters, as produced by likfitLgm

locations

Either a raster, or a single integer giving the number of cells in the X direction which predictions will be made on. If the later the predictions will be a raster of square cells covering the boundin

expPred

Should the predictions be exponentiated, defaults to FALSE.

nugget.in.prediction

If TRUE, predict new observations by adding the nugget effect. The prediction variances will be adjusted accordingly, and the predictions on the natural scale for logged or Box Cox transformed data will be affected. Otherwise predict fitt

Value

• A raster stack is returned with the following layers:
• fixed
  Estimated means from the fixed effects portion of the model
• random
  Predicted random effect
• krige.var
  Conditional variance of predicted random effect (on the transformed scale if applicable)
• predict
  Prediction of the response, sum of fixed and random effects. If exp.pred is TRUE, gives predictions on the exponentiated scale, and half of krige.var is added prior to exponentiating
• predict.log
  If exp.pred=TRUE, the prediction of the logged process.
• predict.boxcox
  If a box cox transformation was used, the prediction of the process on the transformed scale.
• If the prediction locations are different for fixed and random effects (typically coarser for the random effects), a list with two raster stacks is returned.
• prediction
  A raster stack as above, though the random effect prediction is resampled to the same locations as the fixed effects.
• random
  the predictions and conditional variance of the random effects, on the same raster as locations

Details

Given the model parameters and observed data, conditional means and variances of the spatial random field are computed.

See Also

lgm

Examples

data(swissRain)
swissRain$lograin = log(swissRain$rain)
swissRain$elevation = extract(swissAltitude, swissRain)

swissFit = likfitLgm(data=swissRain, 
			trend=lograin~ elevation,
			param=c(range=46500, nugget=0.05,rough=1,  
					aniso.angle.degrees=35, aniso.ratio=12),
			paramToEstimate = c("range","nugget", 
				"aniso.angle.degrees", "aniso.ratio"),
			parscale = c(range=5000,nugget=0.01, 
				aniso.ratio=1,aniso.angle.degrees=5)
)
myTrend = swissFit$trend
myParams = swissFit$param
dput(myParams)
# will give the following
myParams=structure(c(4.49732366951939, 0.000216147764497527, 46575.4897705083, 
0.0918626121620231, 37.3640527963592, 11.9247798102433, 1, 0.686396444665454
), .Names = c("(Intercept)", "elevation", "range", "nugget", 
"aniso.angle.degrees", "aniso.ratio", "rough", "variance"))
myTrend =lograin~ elevation


# make sure krige can cope with missing values!
swissAltitude[1:50,1:50] = NA
swissKrige = krige(data=swissRain, trend = myTrend,
	covariates = swissAltitude,  
	param=myParams,
	locations = 40, expPred=TRUE)



plot(swissKrige[["predict"]], main="predicted rain")
plot(swissBorder, add=TRUE)

# now with box cox and provide a raster for prediction, no covariates
swissFit2 = likfitLgm(data=swissRain, 
			trend=rain~1,
			param=c(range=52000, nugget=0.1,rough=1, boxcox=0.5,
					aniso.angle.degrees=35, aniso.ratio=8),
			paramToEstimate = c("range","nugget", 
				"aniso.angle.degrees", "aniso.ratio"),
			parscale = c(range=5000,nugget=0.01, 
				aniso.ratio=1,aniso.angle.degrees=5)
)
myTrend2 = swissFit2$trend
myParams2 = swissFit2$param
dput(myParams2)
myParams2=structure(c(20.7805835664033, 51161.049007108, 8.64565931800056, 
37.0906650998427, 7.34077366178755, 1, 0.5, 79.3696192640485), .Names = c("(Intercept)", 
"range", "nugget", "aniso.angle.degrees", "aniso.ratio", "rough", 
"boxcox", "variance"))
myTrend2=rain~1

swissRaster = raster(extent(swissBorder), nrows=25, ncols=40)


swissKrige2 = krige(data=swissRain, trend = myTrend2,
	  param=myParams2,
	locations = swissRaster)





plot(swissKrige2[["predict"]], main="predicted rain with box-cox")
plot(swissBorder, add=TRUE)