Beta2parcor: Computation of partial correlation coefficients

Description

This function computes the matrix of partial correlation coefficients
based on the results of the corresponding regression models.

Usage

Beta2parcor(Beta,verbose=FALSE)

Arguments

Beta

matrix of regression coefficients

verbose

print information on conflicting signs etc. Default is verbose=FALSE.

Value

matrix of partial correlation coefficients

Details

A well-known result (Whittaker, 1990) shows that the matrix of
partial correlation coefficients can be estimated by computing a least
squares regression model for each variable. If there are more
variables than observations, the least squares problem is ill-posed
and needs regularization. The matrix Beta stores the regression
coefficients of any user-defined regression method. The function
Beta2parcor computes the
corresponding matrix of partial correlations.

References

J. Whittaker (1990) "Graphical models in applied
multivariate statistics", Wiley, New York.
N. Kraemer, J. Schaefer, A.-L. Boulesteix (2009) "Regularized Estimation of
Large-Scale Gene Regulatory Networks with Gaussian Graphical Models", BMC Bioinformatics, 10:384