Kriging prediction and variance (gstat conventions)

Simple kriging (beta given), universal kriging (X; default a column of ones, ordinary kriging), the GLS trend (blue), block kriging (block from BlockDiscretize or a matrix of equally weighted offsets; nugget left out of the block-block average) and local neighbourhoods (nmax, maxdist), as gstat::krige (Cressie 1993; Wackernagel 2003; Pebesma 2004). Identical to the Python arm morie.fn.krgsys.krige.

Usage

Krige(
  z,
  coords,
  new_coords,
  model,
  X = NULL,
  X0 = NULL,
  beta = NULL,
  block = NULL,
  nmax = NULL,
  maxdist = NULL,
  blue = FALSE
)

Arguments

z: Observations.
coords: Matrix of locations (any dimension).
new_coords: Matrix of targets (block centres with block).
model: Covariance model (KrigingCovariance).
X: Trend design at the data.
X0: Trend design at the targets.
beta: Known trend coefficients (simple kriging).
block: Block discretisation.
nmax: Maximum number of nearest neighbours.
maxdist: Maximum neighbour distance.
blue: Return the GLS trend.

Value

List with prediction, variance, weights, lagrange and, with a global neighbourhood and unknown trend, the GLS beta and trend_residuals.

References

Cressie, N. (1993). Statistics for Spatial Data, rev. edn. Wiley, New York.

Wackernagel, H. (2003). Multivariate Geostatistics, 3rd edn. Springer.

Pebesma, E. J. (2004). Multivariable geostatistics in S: the gstat package. Computers and Geosciences 30, 683-691.

Examples

m <- list(model = "Exp", psill = 1, range = 2, nugget = 0.1)
Krige(c(1, 3, 2), rbind(c(0, 0), c(2, 0), c(0, 2)), rbind(c(1, 1)), m)$prediction
#> [1] 2.060225