acpVIf

Location: CuBatch\Model_PCAVI\acpVIf.m

function [B,PxY,vexpvi,Ayvi,Byvi,vexpX,Ax,Bx] = acpVIf (Y,X,rx,ryvi,algo,optio);

Description:
Calculates the PCA model.If there is no missing data, PCA can be calculated by 'svds' built-in function (algo='Nipals') or by 'svd' built-in function (algo='SVD').

Inputs:
Y: Y data (matrix)
X: X data (matrix)
rx: number of PC for X
ryvi: number of PC for Y
algo: 'SVD' or 'Nipals'
optio: see Matlab help for svds built-in function.

Outputs:
B: regression coefficient matrix
PxY: the projected of Y on [X].
vexpvi: Explained variation of PCA(PxY)
Ayvi: scores of pca(PxY).
Byvi: loadings of pca(PxY).
vexpX: explained variation of PCA(X).
Ax: scores of PCA(X).
Bx: loadings of PCA(X).

Called by:
acpvi

Subroutines:
Internal: None
External: acpf, regression2D

Author:
Christophe Durville
TOTALFINAELF (CReS)
Université Marne-La-Vallée
France

Last modified: 18-Oct-2002 14:03:16

Contact: Christophe Durville, cdurpp@ifrance.com

References