rank.chains: Rank hbfm chains

In mnsekula/hbfm: Generating gene co-expression matrices from scRNA-seq data

Description

Function to rank the MCMC chains by average marginal log-likelihood.

Usage

rank.chains(hbfm.list)

Arguments

hbfm.list

list where each element contains an hbfm.fit-class object; each element of the list contains an object from a different MCMC chain

Details

The rank.chains function is used to rank the input chains based on average marginal log-likelihood in order to identify low performing chains.

Value

data.frame that orders the input hbfm.fit-class objects based on average marginal log-likelihood; the numbers in the chain.num column represent the elements (chains) from hbfm.list

Examples

## Not run: 
## Load dataset
data(gene.mat)

## Run stochastic EM first
## Consider F=5 factors
fit1 <- stoc.em(Y=gene.mat, Fac = 5)

## Run MCMC sampler with initial parameter values from stoc.em
fit.res1 <- hbfm.fit(fit1)

## Run a second chain
fit2 <- stoc.em(Y=gene.mat, Fac = 5, seed = 234)
fit.res2 <- hbfm.fit(fit2, seed = 234)

## Obtain estimated gene-gene correlations from MCMC samples
rank.chains(list(fit.res1, fit.res2))


## End(Not run)

mnsekula/hbfm documentation built on June 29, 2020, 5:12 a.m.