PCG: Preconditioned Conjugate Gradient

In ScoreEB: Score Test Integrated with Empirical Bayes for Association Study

Description

Conduct preconditioned conjugate gradient method to accelerate.

Usage

PCG(G,b,m.marker,sigma.k2,sigma.e2,tol,miter)

Arguments

G	genotype data.
b	column vector.
m.marker	the number of markers.
sigma.k2	variance of polygenic.
sigma.e2	variance of residual error.
tol	convergence threshold.
miter	the maximum number of iterations.

Value

x	x is approximate solution of linear equations.

Examples

data(geno)
G <- t(geno[,-c(1:4)])
n.sample <- dim(G)[1]
m.marker <- dim(G)[2]
b <- rnorm(n.sample)
sigma.k2 <- 6.0
sigma.e2 <- 10.0
tol <- 5e-4
miter <- 20
PCG(G,b,m.marker,sigma.k2,sigma.e2,tol,miter)

ScoreEB documentation built on Sept. 16, 2021, 1:06 a.m.