LORS2: LORS2

In jdrhyne2/FastLORS: Joint Modeling for eQTL Mapping in R

Description

LORS2 is a function used in parameter tuning in LORS. See the parameter tuning section described in Can Yang et al. (2013). This function is adapted from the authors MATLAB implementation

Usage

LORS2(Y, X, L, Omega1, Omega2, B, rho, lambda, tol, maxIter = 1000)

Arguments

Y	gene expression matrix
X	matrix of SNPs
L	matrix of hidden factors
Omega1	Boolean matrix for training data
Omega2	Boolean matrix for validation data
B	a matrix of coefficients for the SNPs
rho	parameter for enforcing sparsity of coefficient matrix
lambda	parameter for enforcing low-rank structure of hidden factor matrix
tol	tolerance level for convergence
maxIter	the maximum number of iterations

Value

B	The estimated coefficients
mu	The estimated intercept
L	The estimated matrix of hidden factors
Err	The residual sum of squares on the validation set

Examples

##Example

#' ## Generate some data
n <- 50
p <- 200
q <- 100
k <- 10
set.seed(123)
X <- matrix(rbinom(n*p,1,0.5),n,p)
L <- matrix(rnorm(n*k),n,k) %*% t(matrix(rnorm(q*k),q,k))
B <- matrix(0, ncol(X), ncol(L))
activeSNPs <- sort(sample(c(1:nrow(B)), 20))
for(i in 1:length(activeSNPs)){
genes_influenced <- sort(sample(c(1:ncol(B)),5))
B[activeSNPs[i], genes_influenced] <- 2
}
E <- matrix(rnorm(n*q),n,q)
Y <- X %*% B + L + E

Omega0 <- !(is.na(Y))
mask <- matrix(runif(nrow(Y)*ncol(Y)) > 0.5, nrow = nrow(Y), ncol = ncol(Y))
Omega1 <- Omega0 & mask
Omega2 <- Omega0 & !mask
rho <- runif(1,3,5)
lambda <- runif(1,3,5)
tol <- 1e-4

## Usage
LORS2(Y, X, L, Omega1, Omega2, B, rho, lambda, tol)

jdrhyne2/FastLORS documentation built on March 5, 2020, 6:50 a.m.