lasich: LASICH estimator

Description Usage Arguments

View source: R/LASICH.R

Description

Calculates sensitivity and specificity of estimating K precision matrices using LASICH. Returns precision matrix estimates, Frobenius norm errors, and operating characteristics.

Usage

1
2
lasich(Sigmas, Sigma0, Omega0, ns, L = NULL, lambda1, lambda2, tol, rho,
  truegraph = NULL, initest = NULL, LD = NULL, wt = NULL, thr = TRUE)

Arguments

Sigmas

(p * p * K) arrays sample covariance matrices from each group

Sigma0

Number of case subjects

Omega0

(p * p * K) arrays true precision matrices

ns

vector of sample sizes for each group

L

graph Laplacian

lambda1

value of first tuning parameter (lambda_1)

lambda2

value of second tuning parameter (lambda_2)

tol

tolerance level for determining convergence of LASICH

rho

tuning parameter for ADMM algorithm

truegraph

a list containing true graphs (to return performance evaluation metrics)

initest

initial estimator for warm start

LD

a diagonal matrix added to L to improve computational efficiency

wt

weight for likelihood; NULL = nk/n, 1 = (1,...,1)???

thr

whether to threshold the sample cov matrix to find connected components of the graph

Sigma0

(p * p * K) arrays true covariance matrices


asondhi/LASICH documentation built on Nov. 24, 2020, 12:36 a.m.