Interface for MOA Stream Clustering Algorithms

Interface for data stream clustering algorithms implemented in the MOA (Massive Online Analysis) framework (Albert Bifet, Geoff Holmes, Richard Kirkby, Bernhard Pfahringer (2010). MOA: Massive Online Analysis, Journal of Machine Learning Research 11: 1601-1604).

CRAN version CRAN RStudio mirror downloads Travis-CI Build Status AppVeyor Build Status

Interface for data stream clustering algorithms implemented in the MOA (Massive Online Analysis) framework. This is an extension package for stream.


  • Stable CRAN version: install from within R.
  • Current development version: Download package from AppVeyor or install via install_git("mhahsler/streamMOA") (needs devtools)


R> library("streamMOA")
# 3 clusters with 5% noise
R> stream <- DSD_Gaussians(k=3, d=2, noise=.05)
# cluster with CluStream  
R> clustream <- DSC_CluStream(m=50, k=3)
R> update(clustream, stream, 500)
R> clustream
CluStream + k-Means (weighted)
Class: DSC_TwoStage, DSC_Macro, DSC 
Number of micro-clusters: 50 
Number of macro-clusters: 3 
# plot micro-clusters
R> plot(clustream, stream, type = "both")

Further Information


Changes in version 1.2-1 (03/20/19)

  • Updated system requirement for new MOA version to Java 8.

Changes in version 1.2-0 (03/14/19)

  • Updated the package to the June 2018 release of the MOA library.
  • Added interfaces to the streamKM, BICO_MOA and DStream_MOA implementations.

Changes in version 1.1-4 (04/06/18)

  • Fixed Java runtime version test for version 10 and beyond.

Changes in version 1.1-3 (01/07/18)

  • Improved documentation for DenStream.
  • Fixed Java SystemRequirement and added runtime version test.

Changes in version 1.1-2 (09/06/15)

  • Interface changes for stream Changes in version 1.2-0.

Changes in version 1.1-1 (07/02/15)

  • Fixed NAMESPACE imports.
  • Downgraded to Java version 1.5 (from 1.7)
  • DSC_DenStream gained parameter k for reclustering.

Changes in version 1.1-0 (05/21/15)

  • updated to MOA Changes in 2014.11
  • DSC_MOA objects can now be saved using the convenience function saveRDS_MOA.
  • DSC_Clustream and DSC_Denstream use now DSC_TwoStage for reclustering.
  • Improved speed by looping through the data points in Java.

Changes in version 1.0-0 (12/18/14)

  • interface with stream changed. We use now update.DSC_MOA, etc.

Changes in version 0.1-1 (6/17/14)

  • applied changes in stream 1.0-2
  • MOA clusterers now report and plot assignment radii.
  • improved vignette

Initial version 0.1-0 (6/10/14)

  • Moved MOA related code to this package
  • updated to MOA Changes in 2014.04

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


1.2-2 by Michael Hahsler, a year ago

Report a bug at

Browse source code at

Authors: Michael Hahsler [aut, cre, cph] , John Forrest [aut, cph] , Matthew Bolanos [ctb] , Matthias Carnein [ctb]

Documentation:   PDF Manual  

GPL-3 license

Imports graphics, stats, methods

Depends on stream, rJava

System requirements: Java (>= 8)

See at CRAN