R functions - Philipp-Neubauer/PopR GitHub Wiki

This is a short overview of R functions - this will likely develop into a proper R package once Julia is more stable (i.e., V>1.0).

Toy_data_creator.R

DPM_call(datas=NULL,baseline=NULL,labels=NULL,learn=FALSE,iters=1000,thin=10,np=1, typeof='N',path.to.julia=getwd(),call_DPM_path=getwd())

implemented in (Julia_call_function.R), calls MCMC for Dirichlet process models
set learn=T for model with baseline
set appropriate path_to_julia if julia files are not in the working directory
typeof: 'N' for normal inference (default, and only one available for now), 'MN' for multinomial (allel frequency data) and 'Both' for a combined analysis

MM_call(datas=NULL,baseline=NULL,labels=NULL,conditional=FALSE,iters=1000,thin=10,np=1, typeof='N',path.to.julia=getwd(),call_DPM_path=getwd())

implemented in (Julia_call_function.R), calls MCMC for finite mixture models
set conditional=T for conditional mixture analysis (e.g., Munch & Clarke 2008), set to False for unconditional (Smith & Campana 2010, Pella & Masusda references).
set appropriate path_to_julia if julia files are not in the working directory
typeof: 'N' for normal inference (default, and only one available for now), 'MN' for multinomial (allel frequency data) and 'Both' for a combined analysis

elink.call(class.ids,path.to.julia=getwd(),elink_path=getwd())

implements exact linkage in Julia
set appropriate path_to_julia if julia files are not in the working directory
class.ids are the individual class.ids as output from the MCMC sampler. Would work with any other sampler as well (e.g., finite mixture models)

as.phylogg(Z,N,tip.label=as.character(1:N))

converts output from exact linkage algorithm (Z) to ape 'phylo' format for tree visualization.
N is the number of tips
tip.label are individual sample labels