TIMBR.approx: Approximate Bayes Factors from TIMBR Output

In wesleycrouse/TIMBR: Tree-based Inference of Multiallelism via Bayesian Regression

Description

Approximate Bayes factors for various hypotheses from TIMBR output

Usage

TIMBR.approx(
  TIMBR.output,
  type = "all",
  ln.ml = F,
  return.prior = F,
  stop.on.error = F
)

Arguments

TIMBR.output	results object from the TIMBR function, prior.M method must be CRP or uniform
type	"all" - report BFs for all allelic series; "merge" - report BFs for biallelic series (i.e. merge analysis); "consistent" - report BFs for consistent merge analysis; prior.M - report BF for list-type prior.M
ln.ml	return ln.ml instead of BF
return.prior	include prior in output
stop.on.error	stop function if error is encountered when using 'integrate'. errors related to roundoff and small values may occur during edge cases

Value

a named vector of approximate Bayes Factors; appoximations that underflow are not reported

Examples

#example data
data(mcv.data)
str(mcv.data)

#call TIMBR using CRP
results <- TIMBR(mcv.data$y, mcv.data$prior.D, mcv.data$prior.M$crp)

#calculate biallelic consistency
TIMBR.approx(results)

wesleycrouse/TIMBR documentation built on Feb. 19, 2021, 7:31 a.m.