compute_phitau: Computes models for the variance components

In CodataGS: Genomic Prediction Using SNP Codata

Description

This function computes the residual variance, the SNP variances and the linear predictor for the SNP variance model.

Usage

compute_phitau(dev, hv, devu, hvu, X.rand.disp)

Arguments

dev	Deviance values
hv	Hat values for the observed response values
devu	Deviance values computed for the random effects
hvu	Hat values for the random effects
X.rand.disp	Design matrix used in the linear predictor for the SNP variance model.

Value

var.e	Residual variance
phi	Vector of SNP variances
coef	Fitted coefficients for the linear predictor in the SNP variance model

Author(s)

Lars Ronnegard

Examples

set.seed(1234)
N <- 20 #Number of individuals
k <- 30 #Number of SNPs with all marker positions including a QTL
#Simulated deviances and hat values
dev <- rnorm(N)^2
hv <- runif(N, 0.1, 0.5)
devu <- rnorm(k)^2
hvu <- runif(k, 0.1, 0.85)
X.rand.disp <- matrix(1, k, 1)
sim.res <- compute_phitau(dev, hv, devu, hvu, X.rand.disp)

CodataGS documentation built on May 17, 2019, 5:03 p.m.