BELA

Bayesian Estimation of Latent Abundance for microbiome data

A Variational Inference suite for NB Ordination Analysis as described in Ren et al (2016).

For MCMC version of the same model, see Boyu Ren's Repo

Documentation is located here

Observations

• In pollution example:
  • Factor is identifiable when sample Y by row, with mean(Y) = log(Theta)
  • It is not identifiable when sample by column, or mean(Y) = Theta