LORS0: LORS0

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

Description

LORS0 is a function for solving the LORS optimization problem through the method described in Can Yang et al. (2013). This function is adapted from the authors MATLAB implementation

Usage

LORS0(Y, X, rho, lambda, maxiter = 1000, eps = 2.2204e-16, tol = 1e-04,
  verbose = FALSE)

Arguments

Y	gene expression matrix
X	matrix of SNPs
rho	parameter for enforcing sparsity of coefficient matrix
lambda	parameter for enforcing low-rank structure of hidden factor matrix
maxiter	maximum number of iterations
eps	constant used when checking the convergence. Ensures no division by 0.
tol	tolerance level for convergence
verbose	chooses whether details should be printed to console. Default is FALSE.

Value

B	The estimated coefficients
mu	The estimated intercept
L	The estimated matrix of hidden factors
f_val_vec	The objective function values
res_vec	The relative change in objective function values

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

rho <- runif(1,3,5)
lambda <- runif(1,3,5)
LORS0(Y, X, rho, lambda)

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