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

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

pop_dtune

calculates the tuning parameters of an orthogonal projection for which the matrix of detrimental information is given

Calling sequence

res_dmt=pop_tune(dmatrix,maxdim,xg,classes_observ,func_reg,x,y,split,lv,(rparam),(centering))

Arguments

dmatrix:

a matrix (n1 x q) or a Div structure containing only detrimental information

maxdim:

the maximum number of dimensions extracted from dmatrix

(xg):

a matrix or a Div structure containing detrimental information

(classes_echant):

a vector identifying the samples in xg

func_reg:

the chosen regression function; a string, ex: 'pls'

x,y:

a calibration dataset: a matrix and a vector or Div structures

split:

for the cross-validation; the number of blocks, or a vector attributing each observation to a CV block

lv:

number of latent variables for the regression

(rparam):

the r parameter; only for the function 'vodka'

(centering):

centred = 1 (by default); not centred=0

res.d_matrix:

the matrix containing only detrimental information

res.d_matrix.d is a matrix of dimensions ((nbr_perturb-1) x q)

res.d_eigenvect:

the eigenvectors of d_matrix

res.d_eigenvect.d is a matrix of dimensions (q x (nbr_perturb-1))

res.d_eigenvalpcent:

the eigenvalues of d_matrix in percent

res.d_eigenvalpcent.d is a vector of dimensions ((nbr_perturb-1) x 1)

res.wilks:

Wilks lambda

res.Wilks.d is a vector of dimensions (nbr_perturb x 1)

res.rmsecv:

rmsecv for several dimensions of orthogonal projection and several of regression

res.rmsecv.d is a matrix of dimensions (lv x nbr_perturb)

res.pls_models:

regression models obtained after an orthogonal projection using 0/1/2/...(nbr_perturb - 1) eigenvectors of res.eigenvect.d

res.pls_models is a list of dimension (nbr_perturb)

help pls for more information about the fields of res.pls_models

Examples

[res]=pop_dtune(d,20,x0, x0_cl_ech,'pls',xcal,ycal,10,5)
[res]=pop_dtune(d,20,x0,x0_cl_ech,'pls',xcal,ycal,10,5,0)
[res]=pop_dtune(d,20,'pls',xcal,ycal,10,5,0)

Authors


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