MCMCglmm (version 2.36)

plotsubspace: Plots covariance matrices

Description

Represents covariance matrices as 3-d ellipsoids using the rgl package. Covariance matrices of dimension greater than 3 are plotted on the subspace defined by the first three eigenvectors.

Usage

plotsubspace(CA, CB=NULL, corr = FALSE, shadeCA = TRUE, 
      shadeCB = TRUE, axes.lab = FALSE, ...)

Arguments

CA

Matrix

CB

Optional second matrix

corr

If TRUE the covariance matrices are transformed into correlation matrices

shadeCA

If TRUE the ellipsoid is solid, if FALSE the ellipsoid is wireframe

shadeCB

If TRUE the ellipsoid is solid, if FALSE the ellipsoid is wireframe

axes.lab

If TRUE the axes are labelled with the eigenvectors

...

further arguments to be passed

Author

Jarrod Hadfield with code taken from the rgl package

Details

The matrix CA is always red, and the matrix CB if given is always blue. The subspace is defined by the first three eigenvectors of CA, and the percentage of variance for each matrix along these three dimensions is given in the plot title.

See Also

rgl

Examples

 if(requireNamespace("rgl")!=FALSE){
   G1<-rIW(diag(4),10)
   G2<-G1*1.2
 #  plotsubspace(G1, G2, shadeCB=FALSE)
 # commented out because of problems with rgl 
 }