graph_invar: Graph Invar

In GraphKit: Estimating Structural Invariants of Graphical Models

Description

Computes the confidence interval for a numeric invariant on an estimated conditional independence graph.

Usage

graph_invar(x, sigmaHat, thetaHat, invar = c("connectivity", "longest_chain",
  "max_degree", "largest_clique", "chromatic_number", "num_singletons",
  "girth"), numB = 1000, alpha = 0.05)

Arguments

x	A sample of data
sigmaHat	A sample covariance matrix for x
thetaHat	A sample precision matrix estimated from sigmaHat
invar	The monotone graph invariant to examine
numB	The number of bootstrap samples to take Default: 1000
alpha	The significance level for the confidence interval Default: 0.05

Value

A confidence intveral for the value of invar with sig. level alpha

Examples

data(Xs,cov.hat,t.hat)
graph_invar(x=Xs,sigmaHat=cov.hat,thetaHat=t.hat,invar="conn")

GraphKit documentation built on May 29, 2017, 5:56 p.m.