path_result_for_roc: The path_result_for_roc() function

In superOmics/sparsenetgls: Using Gaussian graphical structue learning estimation in generalized least squared regression for multivariate normal regression

Description

The path_result_for_roc function is designed to evaluate the the prediction accuracy of a series Gaussian Graphical models (GGM) comparing to the true graph structure. The GGM must use a l-p norm regularizations (p=1,2) with the series of solutions conditional on the regularization parameter.

Usage

path_result_for_roc(PREC_for_graph, OMEGA_path, pathnumber)

Arguments

PREC_for_graph	It is the known precision matrix which is used to assess the estimated precision matrix from GGM.
OMEGA_path	It is a matrix comprising of a series estimated precision matrices from a GGM model using a penalized path based on a range of structure parameters (i.e. λ,\in [0,1]).
pathnumber	It represents the number of graph models (i.e. λ) for the evaluation.The value of pathnumber can be the same number used in a penalized path.

Value

Return the list of assessment results for a series of precision matrices. The results include sensitivity/specificity/NPV/PPV

Examples

prec1 <- matrix(c(0,2,3,1,0,0.5,0,0,0.4),nrow=3,ncol=3)
Omega_est <- array(dim=c(3,3,3))
Omega_est[,,1] <- matrix(c(0,1,2,1,0.5,0.2,0,1,1),nrow=3,ncol=3)
Omega_est[,,2] <- matrix(c(0,1,0,1,0.5,0.2,0,1,1),nrow=3,ncol=3)
Omega_est[,,3] <- matrix(c(0,1,0,1,0,0.2,0,1,1),nrow=3,ncol=3)
rocpath <- path_result_for_roc(PREC_for_graph=prec1,OMEGA_path=Omega_est,
pathnumber=3)

superOmics/sparsenetgls documentation built on Sept. 11, 2020, 5:49 a.m.