SimDirFactorSym: Generate synthetic OTU abundance table from DirFactor model...

In boyuren158/DirFactor: Dependent Dirichlet processes factor model for nonparametric ordination fo microbiome data

Description

Generate synthetic OTU abundance table from DirFactor model with block diagonal-compound symmetric between-sample Gram matrix.

Usage

SimDirFactorSym(vcounts, n, p, m, hyper, K = 2, a = 2, Y.corr = 0.5)

Arguments

vcounts	Required. A vector of total counts per biological sample. Example: vcounts=c(1e5,1e6).
n	Required. Number of biological samples in the simualted table.
p	Required. Number of species in the simulated dataset.
m	Required. True number of factors in the simulated dataset.
hyper	Required. A list of hyper-parameters in the priors. See Details in DirFactor for the fields in this list.
K	Number of diagonal blocks in the Gram matrix of biological samples. Default is 1 (not block diagonal).
a	Random weights of species are generated by sigma*Q^a*(Q>0). Default is 2.
Y.corr	A scalar between -1 to 1. Correlation between biological samples belong to the same block.

Value

A list contains the following fields:

• sigma: a vector with p components of simulated sigma.

• Q: a p*n matrix of normal latent variables.

• X.tru: a m*p matrix of species latent factors.

• Y.tru: a m*n matrix of biological sample latent factors.

• er: a scalar of the simulated pure error.

• data: a list with the same size of vcounts. data[[i]] is the simulated dataset with total counts vcounts[[i]].

Examples

my.hyper = list( nv = 3, a.er = 1, b.er = 0.3, a1 = 3, 
                 a2 = 4, m = 10, alpha = 10, beta = 0 )
SimDirFactorSym( vcounts=c(1e5,1e6), n = 22, p = 68, m = 3, hyper = my.hyper )

boyuren158/DirFactor documentation built on May 13, 2019, 1:38 a.m.