Comments on: Community analysis software

By: Arnaud Browet

Arnaud Browet — Tue, 13 Sep 2011 08:44:01 +0000

Dear Mr Wakita,

Thank you very much for sharing your code with the research community.
I’m working on community detection methods at Université catholique de Louvain (Belgium) and I’m particularly interested in weighted networks.
Is your method able to tackle the community detection problem with weighted adjacency matrix given as input?
If not, do you think it is possible for you to develop this additional option within you java environment.
Thank you very much,

Best Regards,
Arnaud Browet

By: Chris Hamilton

Chris Hamilton — Tue, 19 Jul 2011 04:53:48 +0000

Greetings, I’m running the code for a 10.000 node network with the java runtime memory set to 3 gb (Xmx3g). When I run the analyze command i get a Out Of Memory Error. How much memory I need for such a Network? How much for a one hundred thousand node network?

(I’m running on dual core machine with 4 gb of RAM)

By: Sirius

Sirius — Sat, 02 Jul 2011 15:42:34 +0000

Sorry to bother you again Mr Ken, I have this two questions:

How could I see the modularity of the Network?
AVG K size refers to the average size (number of elements) in every cluster?

By: ken

ken — Fri, 01 Jul 2011 03:42:59 +0000

Great! I will look into your view-clusters problem. Please be waited.

By: sirius

sirius — Thu, 30 Jun 2011 15:39:54 +0000

Hi mr Ken I manage to run version 160. In my network I obtain 13 clusters but when I run the command ‘view-clusters’ it only shows me 3 clusters, why is this?

By: sirius

sirius — Wed, 29 Jun 2011 21:01:21 +0000

Hi Mr Ken, I obtain the following error when I run “java -jar wt2007www.jar –command am2sdb –input karate”:

Exception in thread “main” java.lang.UnsupportedClassVersionError: Bad version number in .class file
at java.lang.ClassLoader.defineClass1(Native Method)
at java.lang.ClassLoader.defineClass(ClassLoader.java:676)
at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:124)
at java.net.URLClassLoader.defineClass(URLClassLoader.java:260)
at java.net.URLClassLoader.access$100(URLClassLoader.java:56)
at java.net.URLClassLoader$1.run(URLClassLoader.java:195)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:188)
at java.lang.ClassLoader.loadClass(ClassLoader.java:317)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:280)
at java.lang.ClassLoader.loadClass(ClassLoader.java:252)
at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:375)

Im running on MacOSX, java version “1.5.0_30″
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_30-b03-389-9M3425)

By: Darine

Darine — Thu, 03 Mar 2011 10:39:39 +0000

Mr Ken,

thank you very very much Sir
and I’ m so sorry for the disturbance.

Darine

By: ken

ken — Thu, 03 Mar 2011 07:42:19 +0000

A newer revision of the software is coming soon.

I packed all the necessary jar files into a single jar file. Now you can run the software simply:

java -jar wt2007www.jar …

You do not have to set up class path and mess around run.sh/run.bat/run.pl scripts.

Added functionality is to dump the clustering process using view-history command.

The software is available from
http://www.is.titech.ac.jp/~wakita/software/wt2007www-161.zip

By: Darine

Darine — Mon, 28 Feb 2011 11:28:40 +0000

Hi Ken

I am working on a research project and wish to compare your community detection algorithm with my algorithm. I want to see that at every step, which two clusters are merging. Using your commands, this does not seem possible. Also you have not included the source code in the zip archive. I wish that you can send me the source code.

Thank you.
Darine

By: Jeff

Jeff — Tue, 25 Jan 2011 17:18:56 +0000

Something simple like this should work.
java -Xmx1g -Djava.ext.dirs=jar jp.ac.titech.is.socialnet.clustering.Main