| predict.se {fields} | R Documentation |
Calculates the standard error of predictions. This is usually the fitted object from a function estimate such as from Krig or Tps.
predict.se(object, ...)
object |
A fitted model object of a certain class |
... |
Additional arguments to be passed to a particular method. e.g. a grid.list or model specification. |
This function is generic and will call the appropriate function to calculate the standard errors for the object class. The prediction standard error is for the estimated function or parameters (a mean value) not for the prediction of a new observation.
A vector of standard errors for the predicted values.
predict, predict.surface.se, predict.se.Krig
fit<-Krig(ozone$x,ozone$y, Covariance="Matern",theta=50, smoothness=1)
predict.se(fit) # std errors of predictions
#
# create a grid of points
xg<- make.surface.grid(
list(East.West=seq(-15,15,,20),North.South=seq(-20,20,,20) ) )
out<- predict.se(fit,xg)
image.plot( as.surface( xg, out))