betahat.app {approximator}

Estimate for beta

Description

Returns the estimator for beta; equation 5. Function betahat.app() returns the estimate in terms of fundamental variables; betahat.app.H() requires the H matrix.

Usage

betahat.app.H(H, V = NULL, Vinv = NULL, z)
betahat.app(D1, subsets, basis, hpa, z, use.Vinv=TRUE)

Arguments

H

In betahat.app.H(), the H matrix, eg that returned by H.fun()

V

Variance matrix

Vinv

Inverse of variance matrix. If not supplied, it is calculated

use.Vinv

In function betahat.app(), a Boolean argument with default TRUE meaning to calculate the inverse of the V matrix; and FALSE meaning to use a method which does not involve calculating the inverse of V. The default method seems to be faster; YMMV

z

vector of observations

D1

Design matrix for level 1 code

subsets

Subsets object

basis

Basis function

hpa

Hyperparameter object

Author(s)

Robin K. S. Hankin

References

M. C. Kennedy and A. O'Hagan 2000. “Predicting the output from a complex computer code when fast approximations are available” Biometrika, 87(1): pp1-13

Examples

data(toyapps)

 betahat.app(D1=D1.toy, subsets=subsets.toy, basis=basis.toy, hpa=hpa.toy, z=z.toy, use.Vinv=TRUE)

 H <- H.fun.app(D1=D1.toy, subsets=subsets.toy, basis=basis.toy,hpa=hpa.toy)
 V <- V.fun.app(D1=D1.toy, subsets=subsets.toy, hpa=hpa.toy)
 betahat.app.H(H=H,V=V,z=z.toy)