pdFactor.pdInd: Factor of a pdInd object.

In gmonette/spida15: Collection of miscellaneous functions for mixed models etc. prepared for SPIDA 2009+ (development version)

Description

Function to compute the upper triangular factor of a pdInd object representing the factorization of the inverse variance matrix.

Usage

pdFactor.pdInd(object)

Arguments

object	a 'pdInd' object from which the right-triangular factor of the variance matrix it represents will be extracted
object	a 'pdInd' object from which the right-triangular factor of the variance matrix it represents will be extracted

Details

Returns a factor for a right log-Cholesky object for positive-definite inverse variance matrix corresponding to a variance matrix with zero covariances except in the first row and column. i.e. $$ V^-1 = R'R $$ with $R$ a right-triangular matrix.

Then if the upper-diagonal elements of $R$ below the first row are all 0 then the corresponding variance matrix with will have zero covariances except on the first row (and column).

Value

the full right-triangular factor, including zeros in the lower triangle, is returned as a vector in column order

the full right-triangular factor, including zeros in the lower triangle, is returned as a vector in column order

Examples

mat <- pdInd(diag(1:4))
pdFactor(mat)
Factor of a pdInd object.

Function to compute the upper triangular factor of a pdInd object
representing the factorization of the inverse variance matrix.

Returns a factor for a right log-Cholesky object for positive-definite
inverse variance matrix corresponding to a variance matrix with zero
covariances except in the first row and column. i.e. $$ V^-1 = t(R)R $$ with
$R$ a right-triangular matrix.

Then if the upper-diagonal elements of $R$ below the first row are all 0
then the corresponding variance matrix with will have zero covariances
except on the first row (and column).

gmonette/spida15 documentation built on May 17, 2019, 7:26 a.m.