VCA: Variance Component Analysis

ANOVA-type estimation (prediction) of random effects in linear mixed models is implemented, following Searle et al. 1991 (ANOVA for unbalanced data). For better performance the SWEEP-Operator is now implemented for generating the ANOVA Type-1 error sum of squares. The primary objective of this package is to perform Variance Component Analyses (VCA). This is a special type of analysis frequently used in verifying the precision performance of diagnostics. The Satterthwaite approximation of the total degrees of freedom and for individual variance components is implemented. These are used in the Chi-Squared tests against a claimed value, and in the respective confidence intervals. Satterthwaite's approximation of denominator degrees of freedom in t-/F-tests of fixed effects are also available. There are several functions for extracting, random effects, fixed effects, variance-covariance matrices of random and fixed effects. Residuals (marginal, conditional) can be extracted as raw, standardized and studentized residuals. Additionally, plotting methods for residuals and random effects and a variability chart are available. The latter is useful for visualizing the variability in sub-classes emerging from the experimental design.

Version: 1.2
Depends: R (≥ 3.0.0)
Imports: stats, graphics, grDevices, Matrix, methods, numDeriv
Published: 2015-07-20
Author: Andre Schuetzenmeister
Maintainer: Andre Schuetzenmeister <andre.schuetzenmeister at roche.com>
License: GPL (≥ 3)
NeedsCompilation: yes
CRAN checks: VCA results

Downloads:

Reference manual: VCA.pdf
Package source: VCA_1.2.tar.gz
Windows binaries: r-devel: VCA_1.2.zip, r-release: VCA_1.2.zip, r-oldrel: VCA_1.2.zip
OS X Snow Leopard binaries: r-release: VCA_1.2.tgz, r-oldrel: VCA_1.1.1.tgz
OS X Mavericks binaries: r-release: VCA_1.2.tgz
Old sources: VCA archive