reverseMAsim.MultiProcess: reverseMAsim.MultiProcess

In SharonLutz/reverseCT: This package contains methods to run mediation analysis and MR with reverse causality

Description

A function to similate the performance of the mediate function from the mediation package in scenarios of reverse causality. Leverages parallel processing and Rcpp with Eigen for faster computation speed.

Usage

reverseMAsim.MultiProcess(n = 1000, pX = 0.2, gamma0 = 0,
  gammaX = 0.1, varM = 1, beta0 = 0, betaX = 1, betaM = c(0, 0.1,
  0.2), varY = 1, nSim = 100, nSimImai = 1000, SEED = 1,
  plot.pdf = T, plot.name = "reverseMAsim.pdf", alpha_level = 0.05,
  use_cpp = F, num_jobs = 2)

Arguments

n	is the sample size.
pX	is the minor allele frequency
gamma0	is the intercept for M
gammaX	is the association of X with M
varM	is the variance of M
beta0	is the intercept for Y
betaX	is the direct effect of X on Y
betaM	is a vector of different associations of M with Y
varY	is the variance of Y
nSim	is the number of simulations to run
nSimImai	is the number of simulations to run in mediate from the mediation package
SEED	is the seed
plot.pdf	is T to output a plot, is F to not output a plot
plot.name	is the name of the plot
alpha_level	is the significance level
use_cpp	use Rcpp with Eigen
num_jobs	the number of cores to use, i.e. the number of parallel procesess to spawn

Value

a matrix of the power of the mediate method from the mediation package to detect an effect of the mediator M on the outcome Y when M and Y are correctly specified and also when they are incorrectly specified (the true mediator is Y and the true outcome is M)

Author(s)

Michael Gooch, Annie Thwing, Sharon Lutz

SharonLutz/reverseCT documentation built on Oct. 30, 2019, 11:54 p.m.