partial_cpc: Partial CPC Algorithm

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

Description

Performs partial CPC Algorithm where the groups share 1 through p-2 CPCs, from Flury (1988) Book. It will return k matrices in a form of a list, showing the common CPC to all matrices and also the uncommon CPC specific to each of the groups. Please ensure that you order the columns of B=new_cpc(matList) according to the first q columns to be kept in this model.

Usage

  partial_cpc(matList, n, q, B = new_cpc(matList, n),
    alpha_stop = 0.001)

Arguments

matList	list of matrices
n	numeric vector of sample sizes
q	the first q components to be kept
B	CPC matrix
alpha_stop	stopping value for small adjustments by jacobi rotations

References

Flury Book 1988

Examples

partial_cpc(test_vole,n_vole,q=1,B=new_cpc(matList,n),alpha_stop=0.001)

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