cvdmngroup: Cross-validation on Dirichlet-Multinomial classifiers.

In DirichletMultinomial: Dirichlet-Multinomial Mixture Model Machine Learning for Microbiome Data

Description

Run cross-validation on Dirichlet-Multinomial generative classifiers.

Usage

cvdmngroup(ncv, count, k, z, ..., verbose = FALSE,
    .lapply = parallel::mclapply)

Arguments

ncv	integer(1) number of cross-validation groups, between 2 and nrow(count).
count	matrix of sample x taxon counts, subsets of which are used for training and cross-validation.
k	named integer() vector of groups and number of Dirichlet components; e.g., c(Lean=1, Obese=3) performs cross-validation for models with k=1 Dirichlet components for the ‘Lean’ group, k=3 Dirichlet components for ‘Obese’.
z	True group assignment.
...	Additional arguments, passed to dmn during each cross-validation.
verbose	logical(1) indicating whether progress should be reported
.lapply	A function used to perform the outer cross-vaildation loop, e.g., lapply for calculation on a single processor, parallel::mclapply for parallel evaluation.

Value

A data.frame summarizing classifications of test samples in cross-validation groups. Columns are:

group	The cross-validation group in which the indivdual was used for testing.
additional columns	Named after classification groups, giving the posterior probability of assignment.

Author(s)

Martin Morgan mailto:[e-mail]

See Also

dmn, DirichletMultinomial-package, vignette("DirichletMultinomial")

Examples

data(xval)  ## result of following commands
head(xval)

## Not run: 
## count matrix
fl <- system.file(package="DirichletMultinomial", "extdata",
                  "Twins.csv")
count <- t(as.matrix(read.csv(fl, row.names=1)))

## phenotype
fl <- system.file(package="DirichletMultinomial", "extdata",
                  "TwinStudy.t")
pheno0 <- scan(fl)
lvls <- c("Lean", "Obese", "Overwt")
pheno <- factor(lvls[pheno0 + 1], levels=lvls)
names(pheno) <- rownames(count)

## subset
keep <- c("Lean", "Obese")
count <- count[pheno 
pheno <- factor(pheno[pheno 

## cross-validation, single Dirichlet component for Lean, 3 for Obese
xval <- cvdmngroup(nrow(count), count, c(Lean=1, Obese=3), pheno,
                   verbose=TRUE, mc.preschedule=FALSE)

## End(Not run)

DirichletMultinomial documentation built on Nov. 8, 2020, 7 p.m.