# krige: Kriging the sample means of statistics In mbaaske/qle: Simulation-Based Quasi-Likelihood Estimation

## Description

`predictKM`,

wrapper for kriging the sample means of statistics

`varKM`,

calculate the kriging prediction variances

`extract`,

extract the results of kriging

## Usage

 ```1 2 3 4 5``` ```predictKM(models, ...) varKM(models, ...) extract(X, type = c("mean", "sigma2", "weights")) ```

## Arguments

 `models` list of covariance models, see `setCovModel` `...` further arguments passed to function `estim` `X` kriging result `type` return type of results, see details below

## Details

For a list of fitted covariance models the function predictKM predicts the sample means of statistics at (unsampled) points, calculates the prediction variances (if '`krig.type`' equals "`var`") at these points and extracts the results. Note that, since we aim on predicting the simulation "error free" value of the sample means, we use a smoothing kriging predictor (see [2, Sec. 3.7.1]).

The function extract either returns the predicted values, the prediction variances or the kriging weights for each point.

## Value

 `predictKM` list of kriging predicted values `varKM` list of kriging prediction variances `extract` matrix of corresponding values (see details)

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12``` ```data(normal) X <- as.matrix(qsd\$qldata[,1:2]) p <- c("mu"=2,"sd"=1) # get simulated statistics at design X Tstat <- qsd\$qldata[grep("^mean.",names(qsd\$qldata))] # predict and extract predictKM(qsd\$covT,p,X,Tstat) # prediction variances varKM(qsd\$covT,p,X,Tstat) ```

mbaaske/qle documentation built on Feb. 3, 2018, 11:02 a.m.