run_bracod: Run the main BRACoD algorithm

In BRACoD.R: BRACoD: Bayesian Regression Analysis of Compositional Data

Run the main BRACoD algorithm

Description

Uses pymc3 to sample the posterior of the model to determine bacteria that are associated with your environmental variable.

Usage

run_bracod(df_relab, env_var, n_sample = 1000, n_burn = 1000, njobs = 4)

Arguments

df_relab	A dataframe of relative microbiome abundances. Samples are rows and bacteria are columns.
env_var	the environmental variable you are evaluating. You need 1 measurement associated with each sample.
n_sample	number of posterior samples.
n_burn	number of burn-in steps before actual sampling stops.
njobs	number of parallel MCMC chains to run.

Value

the pymc trace object which holds the samples of the posterior distribution

Examples

## Not run: 
data(obesity)
r <- simulate_microbiome_counts(obesity)
sim_counts <- r[[1]]
sim_y <- r[[2]]
contributions <- r[[3]]
sim_relab <- scale_counts(sim_counts)
trace <- run_bracod(sim_relab, sim_y, n_sample = 1000, n_burn=1000, njobs=4)

## End(Not run)

BRACoD.R documentation built on March 24, 2022, 5:05 p.m.