tests/meuse_psgp.R

In psgp: Projected Spatial Gaussian Process Methods

library(psgp)
set.seed(100)
# set up data:
data(meuse)
coordinates(meuse) = ~x+y
meuse$value = log(meuse$zinc)
data(meuse.grid)
gridded(meuse.grid) = ~x+y
proj4string(meuse) = CRS("EPSG:28992")
proj4string(meuse.grid) = CRS("EPSG:28992")

# set up intamap object:
psgpObject = createIntamapObject(
  observations = meuse,
  formulaString=as.formula(value~1),
  predictionLocations = meuse.grid,
  class = "psgp"
)

# run test:
checkSetup(psgpObject)

# do interpolation steps:
psgpObject = estimateParameters(psgpObject)

# make prediction
psgpObject = spatialPredict(psgpObject)

# Plot prediction
# plotIntamap(psgpObject)
# plotIntamap(meuse, pch=1, cex=sqrt(meuse$value)/20, add=TRUE)