Learn R Programming

`NPflow`

Overview

NPflow is a package for performing Bayesian estimation of Dirichlet process mixtures of multivariate skew $t$-distributions. It is especially oriented towards flow-cytometry data preprocessing applications, but can be useful for numerous other applications.

The main function of the package is DPMpost().

The method implemented in this package is detailed in the following article:

Hejblum BP, Alkhassim C, Gottardo R, Caron F and Thiebaut R (2019). Sequential Dirichlet Process Mixtures of Multivariate Skew t-distributions for Model-based Clustering of Flow Cytometry Data. The Annals of Applied Statistics, 13(1):638-660. <doi: 10.1214/18-AOAS1209> <arXiv: 1702.04407>

Installation

The easiest way to get NPflow is to install it from CRAN:

install.packages("NPflow")

Or to get the development version from GitHub:

#install.packages("devtools")
devtools::install_github("sistm/NPflow", ref="CRANrelease")

– Boris Hejblum

Copy Link

Version

Install

install.packages('NPflow')

Monthly Downloads

365

Version

0.13.6

License

LGPL-3 | file LICENSE

Maintainer

Boris P. Hejblum

Last Published

August 20th, 2025

Functions in NPflow (0.13.6)

DPMGibbsSkewT_SeqPrior_parallel

Slice Sampling of Dirichlet Process Mixture of skew Student's t-distributions

DPMGibbsSkewT_SeqPrior

Slice Sampling of Dirichlet Process Mixture of skew Student's t-distributions

DPMGibbsSkewT_parallel

Slice Sampling of Dirichlet Process Mixture of skew Student's t-distributions

Posterior estimation for Dirichlet process mixture of multivariate (potentially skew) distributions models

DPMGibbsN_SeqPrior

Slice Sampling of Dirichlet Process Mixture of Gaussian distributions

Slice Sampling of Dirichlet Process Mixture of skew Student's t-distributions

Slice Sampling of the Dirichlet Process Mixture Model with a prior on alpha

DPMGibbsN_parallel

Slice Sampling of the Dirichlet Process Mixture Model with a prior on alpha

DPMGibbsSkewN_parallel

Parallel Implementation of Slice Sampling of Dirichlet Process Mixture of skew normal distributions

Slice Sampling of Dirichlet Process Mixture of skew normal distributions

EM MLE for mixture of NiW

Bayesian Nonparametrics for Automatic Gating of Flow Cytometry data

EM MAP for mixture of sNiW

C++ implementation of the F-measure computation

MLE for sNiW distributed observations

MLE for Gamma distribution

Multiple cost computations with the F-measure as the loss function

EM MLE for mixture of sNiW

Compute a limited F-measure

C++ implementation of the F-measure computation without the reference class 0

cluster_est_pear

Gets a point estimate of the partition using posterior expected adjusted Rand index (PEAR)

cluster_est_Fmeasure

Point estimate of the partition using the F-measure as the cost function.

cluster_est_binder

Point estimate of the partition for the Binder loss function

ELoss of a partition point estimate compared to a gold standard

Scatterplot of flow cytometry data

C++ implementation of similarity matrix computation using pre-computed distances

Burning MCMC iterations from a Dirichlet Process Mixture Model.

cluster_est_Mbinder_norm

Point estimate of the partition using a modified Binder loss function

C++ implementation of multivariate Normal inverse Wishart probability density function for multiple inputs

Probability density function of multiple structured Normal inverse Wishart

print.summaryDPMMclust

Methods for a summary of a DPMMclust object

C++ implementation of multivariate Normal probability density function for multiple inputs

C++ implementation of multivariate skew Normal probability density function for multiple inputs

multivariate Normal inverse Wishart probability density function for multiple inputs

C++ implementation of multivariate structured Normal inverse Wishart probability density function for multiple inputs

C++ implementation of multivariate Normal probability density function for multiple inputs

C++ implementation of multivariate Normal probability density function for multiple inputs

C++ implementation of multivariate Normal probability density function for multiple inputs

Plot of a Dirichlet process mixture of gaussian distribution partition

Sample from a inverse-Wishart distribution

Multivariate log gamma function

multivariate-Normal probability density function

Plot of a Dirichlet process mixture of skew normal distribution partition

multivariate Student's t-distribution probability density function

Convergence diagnostic plots

Computes the co-clustering (or similarity) matrix

C++ implementation

Sample from a Normal inverse-Wishart distribution whose parameter are given by the structure hyper

C++ implementation of multivariate normal probability density function for multiple inputs

Construction of an Empirical based prior

Generating cluster data from the Chinese Restaurant Process

Sample from a normal inverse Wishart distribution whose parameter are given by the structure SufStat

Plot of a Dirichlet process mixture of skew t-distribution partition

multivariate Skew-Normal probability density function

C++ implementation of multivariate skew normal likelihood function for multiple inputs

postProcess.DPMMclust

Post-processing Dirichlet Process Mixture Models results to get a mixture distribution of the posterior locations

multivariate skew-t probability density function

C++ implementation of multivariate skew t likelihood function for multiple inputs

summary.DPMMclust

Summarizing Dirichlet Process Mixture Models

Sampler for the concentration parameter of a Dirichlet process

similarityMat_nocostC

C++ implementation

C++ implementation of the multinomial sampling from a matrix of column vectors, each containing the sampling probabilities for their respective draw

Sample from a Wishart distribution

vclust2mcoclustC

C++ implementation

C++ implementation of residual trace computation step used when sampling the scale

Return updated sufficient statistics S with new data matrix z

Return updated sufficient statistics S for skew t-distribution with data matrix z

Return updated sufficient statistics S with new data matrix z