Package: VCA
Version: 1.3
Date: 2016-03-23
Title: Variance Component Analysis
Author: Andre Schuetzenmeister <andre.schuetzenmeister@roche.com>
Maintainer: Andre Schuetzenmeister <andre.schuetzenmeister@roche.com>
Depends: R (>= 3.0.0)
Imports: stats, graphics, grDevices, lme4, Matrix, methods, numDeriv
Description: 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 implemented
             for generating the ANOVA Type-1 error sum of squares. Restricted Maximum Likelihood (REML) estimation is 
             also available making use of the 'lme4' package.
             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 for models fitted by ANOVA as well as for models fitted by REML. 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. For models fitted by (REML) the variance-covariance matrix of the variance components is approximated
             by the method pointed out in Giesbrecht & Burns (1985), which is required for applying the Satterthwaite approximation
             of the degrees of freedom. For models fitted by ANOVA the method pointed out in Searle et al. (1991) employing quadratic
             forms generating ANOVA sum of squares is 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.
License: GPL (>= 3)
Collate: 'plot.R' 'vca.R'
NeedsCompilation: yes
Packaged: 2016-03-23 09:08:16 UTC; schueta6
Repository: CRAN
Date/Publication: 2016-03-23 23:41:04
