Calculate Variance-Covariance Matrix of Variance Components of 'VCA' objects.

Share:

Description

This function computes the variance-covariance matrix of variance components (VC) either applying the approach given in the 1st reference ('method="scm"') or using the approximation given in the 2nd reference ('method="gb"').

Usage

1
vcovVC(obj, method = NULL, quiet = FALSE)

Arguments

obj

(VCA) object

method

(character) string, optionally specifying whether to use the algorithm given in the 1st reference ("scm") or in the 2nd refernce ("gb"). If not not supplied, the option is used coming with the 'VCA' object.

quiet

(logical) TRUE = will suppress any warning, which will be issued otherwise

Details

When 'method="scm"' is used function getVCvar is called implementing this rather time-consuming algorithm. Both approaches, respectively the results they generate, diverge for increasing degree of unbalancedness. For balanced designs, they seem to differ only due to numerical reasons (error propagation).

This function is called on a 'VCA' object, which can be the sole argument. In this case the value assigned to element 'VarVC.method' of the 'VCA' object will be used (see getVCvar for computational details).

Value

(matrix) corresponding to variance-covariance matrix of variance components

Author(s)

Andre Schuetzenmeister andre.schuetzenmeister@roche.com

References

Searle, S.R, Casella, G., McCulloch, C.E. (1992), Variance Components, Wiley New York

Giesbrecht, F.G. and Burns, J.C. (1985), Two-Stage Analysis Based on a Mixed Model: Large-Sample Asymptotic Theory and Small-Sample Simulation Results, Biometrics 41, p. 477-486

Examples

1
2
3
4
5
6
7
8
9
## Not run: 
data(realData)
dat1 <- realData[realData$PID==1,]
fit  <- anovaVCA(y~lot/calibration/day/run, dat1, SSQ.method="qf") 
vcovVC(fit)
vcovVC(fit, "scm")		# Searle-Casella-McCulloch method (1st reference)
vcovVC(fit, "gb")		# Giesbrecht and Burns method (2nd reference)

## End(Not run)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.