mvrnorm_corr_gen: Generate Multivariate Random Normal Longitudinal Data

In microbiomeDASim: Microbiome Differential Abundance Simulation

Description

For this methodology we assume that we draw a set of n independent each with q_{i} observations.

Usage

mvrnorm_corr_gen(
  n,
  obs,
  t,
  mu,
  sigma,
  rho,
  corr_str = c("ar1", "compound", "ind"),
  zero_trunc = TRUE
)

Arguments

n	integer scalar representing the total number of individuals
obs	vector of length n specifying the number of observations per indivdiual.
t	vector corresponding to the timepoints for each individual.
mu	vector specifying the mean value for individuals.
sigma	scalar specifying the standard deviation for all observations.
rho	numeric scalar value between [0, 1] specifying the amount of correlation between. assumes that the correlation is consistent for all subjects.
corr_str	character value specifying the correlation structure. Currently available methods are \'ar1\', \'compound\', and \'ind\' which correspond to first-order autoregressive, compound or equicorrelation, and independence respecitvely.
zero_trunc	logical value to specifying whether the generating distribution should come from a multivariate zero truncated normal or an untruncated multivariate normal. by default we assume that zero truncation occurs since this is assummed in our microbiome setting.

Value

This function returns a list with the following objects:

df - data.frame object with complete outcome Y, subject ID, time, group, and outcome with missing data

Y - vector of complete outcome

Mu - vector of complete mean specifications used during simulation

Sigma - block diagonal symmetric matrix of complete data used during simulation

N - total number of observations

Examples

size <- 15
reps <- 4
N <- size*reps
mvrnorm_corr_gen(n=size, obs=rep(reps, size), t=rep(seq_len(4), size),
mu=rep(1, N), sigma=2, rho=0.9, corr_str="ar1")

microbiomeDASim documentation built on Nov. 8, 2020, 10:58 p.m.