bp.error: calculate back-projection errors

In bbolker/cpcbp: common principal components/back-projection analysis

Description

Calculates the size of back-projection errors, based on errors of common principal components

Usage

bp.error(x,f,cpcmat, m, eigvar,
use="complete.obs",debug=FALSE,
center=TRUE)

Arguments

x	a numeric matrix (or data frame with all numeric values, or (if f is missing) a list of a data matrix and a grouping variable
f	a factor describing the group structure of the data
cpcmat	common principal components matrix (if already computed)
m	matrix of mean values of variables by group (if already computed)
eigvar	variances of eigenvalues (if already computed)
use	method for missing observations when computing covariances (see cov for details)
debug	(logical) print debugging information?
center	center group means on grand mean?

Details

Uses CPC, variances of CPC (eigenvector) components, and means of groups to calculate total back-projection errors in variables. If cpcmat is not provided, calls cpcvecfun to compute CPC; if eigvar is not provided, calls calc.cpcerr to calculate variances on eigenvalues.

Value

If new=TRUE, a matrix of back-projection errors by variable and group; otherwise, a vector of total back-projection errors in each variable.

Author(s)

Ben Bolker

References

Flury

Examples

   set.seed(1001)
   X <- simdata(npts=100)
   bp.error(X)

bbolker/cpcbp documentation built on May 11, 2019, 9:28 p.m.