dppMCMC_C {DPP}

A Reference Class that provides DPP functionality

Description

This class implements the main functionality of this package. The consturctor receives a numeric vector (Y) and priors. A model should be provided specifying the distributions to be used for inference (e.g. NormalModel for Normal distributions or GammaModel for Gamma distributions). Then an MCMC algorithm will be used to infer a number of distributions (k) that fit the data. The prior for the number of distributions is specified by the concentration_parameter_alpha and expected_k. Once the data and priors are specified the method run is used to start the inference.

Details

Class dppMCMC_C Class dppMCMC_C A Reference Class that provides DPP functionality

Fields

dpp_mcmc_object

a DPPmcmc object

Methods

getNumCategoryProbabilities(burnin_cutoff = 0.25)

returns the probabilities vector for inferred number of categories

getNumCategoryTrace(burnin_cutoff = 0.25)

returns the trace vector for the inferred number of categories in the data

initialize(data, output, model, num_auxiliary_tables = 4, expected_k = 2, power = 1, verbose = TRUE, sample.num.clusters = TRUE)

the class constructor, initializes DPPmcmc object with data and parameters

run(generations, sample_freq = generations/1000, log_file, allocation_file, param_file, append = TRUE, random = FALSE, auto_stop = FALSE, min_ess = 500, max_gen = 1e+05)

starts the MCMC run

Examples

normal.model<-new(NormalModel,
                 mean_prior_mean=0.5,
                 mean_prior_sd=0.1,
                 sd_prior_shape=3,
                 sd_prior_rate=20,
                 estimate_concentration_parameter=TRUE,
                 concentration_parameter_alpha=10,
                  proposal_disturbance_sd=0.1)

#simulating three normal distributions
y <- c(rnorm(100,mean=0.2,sd=0.05), rnorm(100,0.7,0.05), rnorm(100,1.3,0.1))
hist(y,breaks=30)

#setwd("~/yourwd") #mcmc log files will be saved here
## Not run: 
my_dpp_analysis <- dppMCMC_C(data=y,
                             output = "output_prefix_",
                             model=normal.model,
                             num_auxiliary_tables=4,
                             expected_k=1.5,
                             power=1)
#running the mcmc  , generations will be ignored because auto_stop=true
my_dpp_analysis$run(generations=1000,auto_stop=TRUE,max_gen = 10000,min_ess = 500)

#we get rid of the first 25% of the output (burn-in)
hist(my_dpp_analysis$getNumCategoryTrace(0.25))

my_dpp_analysis$getNumCategoryProbabilities(0.25)

## End(Not run)