Clusterer – Analyze Network Cluster With Graphical Visualization


Computers are usually invented as single machines, but we can also network them together to create a computer cluster. A computer cluster is a chain of computers working together to perform like a single computer. Clusters are usually used to improve performance of a network or service. It is different from load balancing, as clustering is the use of multiple computers to provide a single service, whereas, load balancing is a technique to use multiple computers in a cluster. If you have formed a cluster, it can be difficult to analyze it’s ability and to obtain data. Clusterer is a Java-based portable application which provides analysis and visualized network graphs to analyze network clusters using various algorithms.

The main interface of Clusterer provides numerous algorithms to perform network cluster analysis. Select an Algorithm from the drop down menu and click Analyze to obtain data. The available algorithms include;

  • Pons-Latapy Random Walk t=log (n)
  • Pons-Latapy Random Walk t=log (n)+1
  • Pons-Latapy Random Walk t=log (n)-1
  • Gerwan Newman with Strong Community
  • Gerwan Newman with Weak Community
  • Wishart-MacQueen
  • K-Means
  • MacNaughton-Smith

The use of any of the above algorithms will depend on your cluster type. For example, the Girvan–Newman algorithm is used to detect communities in complex systems. A community consists of a subset of nodes. You can also use the Generate option to check cluster arrangement and to obtain data count.


Once you click Analyze, you will be able to obtain information from the available options. For example, you can click on GFX to receive an index cluster graph. Since Clusterer contains many numerical calculations, it is more useful for administrators handling large clusters, with in depth information about network analysis algorithms rather than admins running small clusters with little experience and knowledge of network cluster analysis methods.

Clusterer GFX

You can view alternative cluster arrangements from the View Alternatives option. The data obtained from the overall analysis can also be saved to a text file for later review from File –> Save.


If you are running multiple clusters, you can get a graph view for the largest cluster using the option ” Get the biggest cluster as a graph” from the main interface. Other options serve the following purposes:

  • Perturb: can be used to disrupt the distance in data for calculation purposes
  • Correct: corrects the data edge length to the shortest path
  • DataSet: generates a data set
  • Graph: generates data graphs

Clusterer is an  open source application which works on Windows XP, Windows Vista and Windows 7.

Download Clusterer