GSproc {GPoM}

Gram-Schmidt procedure

Description

Computes regressors coefficients using the Gram-Schmidt procedure.

Usage

GSproc(polyK, ivec, weight = NULL)

Arguments

polyK

One list including $Y and $phy with: $Y a matrix for which the ith column will be used to add one orthogonal vector to the (i-1)th vectors of the current orthogonal base; and $phy such as the current orthogonal base is given by the (i-1)th first columns of matrix polyK$phy.

ivec

Defines i, the current vector of polyK$Y and the current orthogonal base of pParam$phy.

weight

The weighing vector.

Value

uNew The model parameterization, that is: The residual orthogonal vector that can be included into the current orthogonal base. If the current base is empty, uNew is equal to the input vector of $Y; if the base is complete, uNew equals 0.

Author(s)

Sylvain Mangiarotti