Learn R Programming

⚠️There's a newer version (0.4.2) of this package.Take me there.

dirichletprocess

The dirichletprocess package provides tools for you to build custom Dirichlet process mixture models. You can use the pre-built Normal/Weibull/Beta distributions or create your own following the instructions in the vignette. In as little as four lines of code you can be modelling your data non-parametrically.

Installation

You can install dirichletprocess from github with:

# install.packages("devtools")
devtools::install_github("dm13450/dirichletprocess")

For a full guide to the package and its capabilities please consult the vignette:

browseVignettes(package = "dirichletprocess")

Examples

Density Estimation

Dirichlet processes can be used for non-parametric density estimation.

faithfulTransformed <- faithful$waiting - mean(faithful$waiting)
faithfulTransformed <- faithfulTransformed/sd(faithful$waiting)
dp <- DirichletProcessGaussian(faithfulTransformed)
dp <- Fit(dp, 100, progressBar = FALSE)
plot(dp)

data.frame(Weight=dp$weights, Mean=c(dp$clusterParameters[[1]]), SD=c(dp$clusterParameters[[1]]))
#>        Weight       Mean         SD
#> 1 0.371323529 -1.1756510 -1.1756510
#> 2 0.625000000  0.6597522  0.6597522
#> 3 0.003676471  0.1061095  0.1061095

Clustering

Dirichlet processes can also be used to cluster data based on their common distribution parameters.

faithfulTrans <- as.matrix(apply(faithful, 2, function(x) (x-mean(x))/sd(x)))
dpCluster <-  DirichletProcessMvnormal(faithfulTrans)
dpCluster <- Fit(dpCluster, 1000, progressBar = FALSE)

To plot the results we take the cluster labels contained in the dp object and assign them a colour

For more detailed explanations and examples see the vignette.

Copy Link

Version

Install

install.packages('dirichletprocess')

Monthly Downloads

441

Version

0.2.0

License

GPL-3

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/dm13450/dirichletprocess

Maintainer

Dean Markwick

Last Published

January 29th, 2018

Functions in dirichletprocess (0.2.0)

BetaMixtureCreate

Create a Beta mixing distribution.

ChangeObservations

Change the observations of fitted Dirichlet Process.

DirichletProcessHierarchicalBeta

Create a Hierarchical Dirichlet Mixture of Beta Distributions

DirichletProcessMvnormal

Create a Dirichlet mixture of multivariate normal distributions.

GaussianMixtureCreate

Create a Normal mixing distribution

GlobalParameterUpdate

Update the parameters of the hierarchical Dirichlet process object.

Draw from the prior distribution

PriorParametersUpdate

Update the prior parameters of a mixing distribution

DirichletProcessCreate

Create a Dirichlet Process object

DirichletProcessGaussian

Create a Dirichlet Mixture of Gaussians

PosteriorClusters

Generate the posterior clusters of a Dirichlet Process

Draw from the posterior distribution

DirichletProcessWeibull

Create a Dirichlet Mixture of the Weibull distribution

Fit the Dirichlet process object

PosteriorFunction

Generate the posterior function of the Dirichlet function

PosteriorParameters

Calculate the posterior parameters for a conjugate prior.

WeibullMixtureCreate

Create a Weibull mixing distribution.

ClusterParameterUpdate

Update the cluster parameters of the Dirichlet process.

DirichletProcessBeta

Dirichlet process mixture of the Beta distribution.

MixingDistribution

Create a mixing distribution object

ClusterComponentUpdate

Update the component of the Dirichlet process

ClusterLabelPredict

Predict the cluster labels of some new data.

Mixing Distribution Likelihood

Create a multivariate normal mixing distribution

The Stick Breaking representation of the Dirichlet process.

Update the Dirichlet process concentration parameter.

plot.dirichletprocess

Plot the Dirichlet Process

LikelihoodFunction

The Likelihood of a Dirichlet process object.

Calculate how well the prior predicts the data.

Calculate the prior density of a mixing distribution

HierarchicalBetaCreate

Create a Mixing Object for a hierarchical Beta Dirichlet process object.

Initialise a Dirichlet process object

weighted_function_generator

Generate a weighted function.

dirichletprocess

A flexible package for fitting Bayesian non-parametric models.

Tumour incidences in rats