<< pop_dextract (X->X) Pretraitments by orthogonal projection pop_dtune >>

fact >> (X->X) Pretraitments by orthogonal projection > pop_dtune

pop_dtune

calculates the tuning parameters of the TOP-Transfer by Orthogonal Projection method, for which the variability is issued from two measurements on the same samples (ex: with two devices)

Calling sequence

res=pop_dtop_optim(x,y,xtest1,xtest2,maxdim,lv)

Arguments

x,y:

a calibration dataset; x and y are of dimensions (n x q) and (n x 1) respectively; datasets or div structures

xtest1,xtest2:

two measurements on the same samples; two matrices (n0 x q) or div structures

maxdim:

the max number of dimensions to remove; an integer

lv:

the number of latent variables for the PLS regression; an integer

res:

a structure contaning the following fields:

- res.d_matrix: the matrix D = xtest1 - xtest2, of dimensions (n0 x q)

- res.eigenvec: the eigenvectors of D; a matrix (q x n1)

- res.ev_pcent: a vector (n1 x 1)

- res.wilks: the Wilks lambda calculated on D after deflation of 0 to maxdim dimensions; a vector (maxdim+1,1)

- res.rmsecv: the errors of prediction in cross validation; a matrix (lv x (maxdim+1))

Examples

[res]=pop_dtop_optim(x,y,xtest1,xtest2,5,12)

Authors


Report an issue
<< pop_dextract (X->X) Pretraitments by orthogonal projection pop_dtune >>