FacOptScale

Location: ..\cubatch\Model_PARAFAC\private\FacOptScale.m

function [As,Bs,Cs,...,as,bs,cs] = FacOptScale (A,B,C,...,a,b,c,...);

Description:
Scales the loading vectors of each factors in the two sets so that the euclidean distance between the two of them is minimised:

The set indicated with a_m (the loading matrix for the m-th mode) is used as a reference.
It uses a constrained optimisation routine (via the function parafacscaerr) and the scaling factor has to be between 0 and 2.
The norm of each factor is equally spread over the vectors defining it:


Inputs:
A,B,C,...: loadings set to be scaled
a,b,c,...: reference set

Outputs:
As,Bs,Cs,...: optimally scaled loadings set
as,bs,cs,...: scaled reference set
 

Called by:
Model_PARAFAC\Compute_Risk

Subroutines:
Internal: isodd (returns 1 if argument is odd), normvec (normalises vector to length 1)
External: parafacscaerr, scale_factors

Author:
Giorgio Tomasi
Royal Agricultural and Veterinary University
MLI, LMT, Chemometrics group
Rolighedsvej 30
DK-1958 Frederiksberg C
Danmark

Last modified: 11-Oct-2002 19:43:28

Contact: Giorgio Tomasi, gt@kvl.dk

References
[1] - Riu J. and Bro R., "Jack knife for estimation of standard errors and outlier detection in PARAFAC models", Chemom.Intell.Lab.Syst., Vol. 65, 2003, 35-49
[2] - Tomasi G. and Bro R. "Fitting the PARAFAC model", Comp.Stat.Data Anal., 2002, Submitted