simulate_DM: Title: Simulate Dirichlet-Multinomial Regression Model

In Goodgolden/LDTM: Longitudinal Dirichlet Tree Multinomial Model

Title: Simulate Dirichlet-Multinomial Regression Model

Description

Title: Simulate Dirichlet-Multinomial Regression Model

Usage

simulate_DM(
  n_obs = 100,
  n_vars = 30,
  n_taxa = 75,
  n_relevant_vars = 4,
  n_relevant_taxa = 4,
  beta_min = 1,
  beta_max = 1.25,
  signoise = 1,
  n_reads_min = 5000,
  n_reads_max = 10000,
  theta0 = 0.01,
  rho = NULL,
  Sigma = NULL
)

Arguments

n_obs	integer Number of observations
n_vars	integer Number of variables
n_taxa	integer Number of taxa
n_relevant_vars	integer Number of variables that are important
n_relevant_taxa	integer Number of important taxa
beta_min	numeric The desired min beta value
beta_max	numeric The desired max beta value
signoise	numeric Default is 1.
n_reads_min	numeric Default is 5000. Used to obtain a random subject from n_reads_min to n_reads_max from the Dirichlet Multinomial Distribution using dirmult::simPop()
n_reads_max	numeric Default is 10000. Used to obtain a random subject from n_reads_min to n_reads_max from the Dirichlet Multinomial Distribution using dirmult::simPop()
rho	numeric Used for covariate correlation between 0 & 1. Default is NULL. Need either rho or Sigma to be input.
Sigma	matrix A symmetric positive definite matrix used for covariate correlation structure columns must match number of covariates. Default is NULL. Need either rho or Sigma to be input.

Value

A list of outcomes including:

• X = XX

• Y = YY

• alphas = intercept,

• betas = betas,

• n_reads_min = n_reads_min,

• n_reads_max = n_reads_max,

• theta0 = theta0,

• phi = phi,

• rho = rho,

• signoise = signoise,

• Sigma = Sigma

This function can be used to simulate DM data.

Taken from Wadsworth (2017) An integrative Bayesian Dirichlet-Multinomial regression model for the analysis of taxonomic abundances in microbiome data

Goodgolden/LDTM documentation built on May 25, 2022, 5:25 p.m.