predict.geolm_cmodStd function

Predict method for geostatistical models

predict calculates the predicted values at specified locations. The method can additionally provide the mean square prediction error (mspe) and perform conditional simulation.

## S3 method for class 'geolm_cmodStd'
predict(
  object,
  newdata,
  nsim = 0,
  return_type = "SpatialPointsDataFrame",
  dmethod = "chol",
  compute_mspe = TRUE,
  sp = NULL,
  ...
)

Arguments

• object: An object produced by the geolm

  function.

• newdata: An optional data frame in which to look for the coordinates at which to predict. If omitted, the observed data locations are used.

• nsim: A non-negative integer indicating the number of realizations to sample at the specified coordinates using conditional simulation.

• return_type: A character string indicating the type of object that should be returned. The default is "SpatialPointsDataFrame" for easy plotting of results (see Examples). Other options include "data.frame", "geardf", and "sf".

• dmethod: The method used to decompose the covariance matrix for conditional simulation. Valid options are "chol", "eigen", and "svd". The default is "chol".

• compute_mspe: A logical value indicating whether the mean square prediction error should be calculated. Default is TRUE.

• sp: This argument will be deprecated in the future. Please use the return_type argument. A logical value indicating whether to object returned should be of class SpatialPointsDataFrame for easier plotting with the sp package. Default is NULL.

• ...: Currently unimplemented.

Returns

A data.frame, SpatialPointsDataFrame, geardf, or sf

object with the kriging predictions pred, kriging variance/mean-square prediction error (mspe), the root mean-square prediction error mspe (rmspe), and the conditional simulations sim.1, sim.2, etc. sim.1, sim.2, etc.

Details

The newdata data frame must include the relevant covariates for the prediction locations, where the covariates are specified on the right side of the ~ in object$formula. newdata must also include the coordinates of the prediction locations, with these columns having the names provided in object$coordnames.

Examples

# generate response
y = rnorm(10)
# generate coordinates
x1 = runif(10); x2 = runif(10)

# data frame for observed data
data = data.frame(y, x1, x2)
# newdata must have columns with prediction coordinates
newdata = data.frame(x1 = runif(5), x2 = runif(5))

# specify a standard covariance model
mod = cmod_std(model = "exponential", psill = 1, r = 1)

# geolm for universal kriging
gearmod_uk = geolm(y ~ x1 + x2, data = data, mod = mod,
                   coordnames = c("x1", "x2"))
# prediction for universal kriging, with conditional simulation
pred_uk = predict(gearmod_uk, newdata, nsim = 2)

# demonstrate plotting abilities if return_type == "geardf"
 pred_geardf = predict(gearmod_uk, newdata,
               return_type = "geardf")
 plot(pred_geardf, "pred")
 plot(pred_geardf, interp = TRUE)

# demonstrate plotting abilities if sp package installed
if (requireNamespace("sp", quietly = TRUE)) {
 pred_spdf = predict(gearmod_uk, newdata,
             return_type = "SpatialPointsDataFrame")
 sp::spplot(pred_spdf, "pred")
 sp::spplot(pred_spdf)
}
# demonstrate plotting abilities if sf package installed
if (requireNamespace("sf", quietly = TRUE)) {
 pred_sfdf = predict(gearmod_uk, newdata,
             return_type = "sf")
 plot(pred_sfdf["pred"])
 plot(pred_sfdf)
}

# geolm for ordinary kriging
gearmod_ok = geolm(y ~ 1, data = data, mod = mod,
                   coordnames = c("x1", "x2"))
# prediction for ordinary kriging
pred_ok = predict(gearmod_ok, newdata)

# geolm for simple kriging
gearmod_sk = geolm(y ~ 1, data = data, mod = mod,
                 coordnames = c("x1", "x2"), mu = 1)
# prediction for simple kriging
pred_sk = predict(gearmod_sk, newdata)

Author(s)

Joshua French