param-utils: Functions for the mapping between spherical and Euclidean...

In graphpcor: Models for Correlation Matrices Based on Graphs

Functions for the mapping between spherical and Euclidean coordinates.

Description

Functions for the mapping between spherical and Euclidean coordinates.

Usage

rphi2x(rphi)

x2rphi(x)

rtheta(n, lambda = 1, R, theta.base)

dtheta(theta, lambda, theta.base, H.elements)

KLD10(C1, C0)

theta2H(theta)

Arguments

rphi	numeric vector where the first element is the radius and the remaining are the angles
x	parameters in the Euclidian space to be converted
n	integer to define the size of the correlation matrix
lambda	numeric as the parameter for the Exponential distribution of the radius
R	scaling matrix (square root of the Hessian around the base model)
theta.base	numeric vector of the base model
theta	numeric vector of length m.
H.elements	list output of theta2H
C1	is a correlation matrix.
C0	is a correlation matrix of the base model.

Details

see N-sphere/Euclidian

compute C1 using 'theta2C' on theta with

KLD = 0.5( tr(C0^{-1}C1) -p + ... - log(|C1|) + log(|C0|) )

Functions

• x2rphi(): Tranform from Euclidian coordinates to spherical

• rtheta(): Drawn samples from the PC-prior for correlation

• dtheta(): PC-prior density for the correlation matrix

• KLD10(): Compute the KLD with respect to a base model

• theta2H(): Compute the hessian, its svd and some elements

graphpcor documentation built on June 8, 2025, 10:37 a.m.