JNICoCo - Java API for CoCoNAD

Download

Description

JNICoCo is an Java Application Programmers Interface (API) for the CoCoNAD implementation in C that also includes functions for generating a pattern spectrum with the help of surrogate data and CoCoNAD analysis and for pattern spectrum estimation. The library underlying this API was implemented using the Java Native Interface (JNI).

If you have trouble getting the library to work on Microsoft Windows, check whether you have the Microsoft Visual C++ Redistributable for Visual Studio 2022 (see under "Other Tools and Frameworks") installed, as the library was compiled with Microsoft Visual Studio 2022.

Installation

On Microsoft Windows it should suffice to place the dynamic link library (JNICoCo.dll) into the same directory from which the Java class or the Java archive are called that are meant to use the API provided by the library. On a Unix/Linux system it may also suffice to place the shared object (libJNICoCo.so) into such a directory, but this is less likely to work.

If this simple method fails, the reason is usually that Java cannot find the library, due to the fact that the necessary path is missing from the Java library path. To fix this, the Java class or Java archive that is meant to use the library should be invoked from the command line with

java [-jar] -Djava.library.path="." <path_to_class_or_jar>

Here the option -Djava.library.path="." adds the current working directory (denoted by ".") to the list of paths that are searched by Java for the library libJNICoCo.so (GNU/Linux) or JNICoCo.dll (Microsoft Windows). Note that the library may also be placed into any other directory if the "." is replaced by the path to that directory. Note also that the invocation command may conveniently be stored in a batch file (for Microsoft Windows) or a shell script (for GNU/Linux), thus hiding the unusual invocation from an end user.

In order to develop an application based on this API, the Java sources made available above are needed. The main file is coconad/JNICoCo.java, which contains the functions through which the C implementation can be accessed. The other Java source files provide classes for (parallel) trains and pattern sets. See the Java documentation that accompanies the sources for details.

References

More information about the CoCoNAD algorithm can be found in the following papers:

• Frequent Item Set Mining for Sequential Data: Synchrony in Neuronal Spike Trains
  David Picado-Muiño and Christian Borgelt
  Intelligent Data Analysis 18(6):997-1012.
  IOS Press, Amsterdam, Netherlands 2014
  doi:10.3233/IDA-140681 iospress.com
  (16 pages)
• Finding Frequent Patterns in Parallel Point Processes
  Christian Borgelt and David Picado-Muiño
  Proc. 12th Int. Symposium on Intelligent Data Analysis, 116-126.
  Springer-Verlag, Berlin/Heidelberg, Germany 2013
  ida_13.pdf (443 kb) ida_13.ps.gz (491 kb) (11 pages)

A description of the methods of pattern spectrum filtering and pattern set reduction can be found in the following papers:

• Finding Neural Assemblies with Frequent Item Set Mining
  David Picado-Muiño, Christian Borgelt, Denise Berger, George Gerstein, and Sonja Grün
  Frontiers in Neuroinformatics 7:article 9
  Frontiers Media, Lausanne, Switzerland 2013
  doi:10.3389/fninf.2013.00009 frontiersin.org
  accfim.pdf (1797 kb) accfim.ps.gz (772 kb) (14 pages)
• Statistical Evaluation of Synchronous Spike Patterns Extracted by Frequent Itemset Mining
  Emiliano Torre, David Picado-Muiño, Michael Denker, Christian Borgelt, and Sonja Grün
  Frontiers in Computational Neuroscience, 7:article 132
  Frontiers Media, Lausanne, Switzerland 2013
  doi:10.3389/fncom.2013.00132 frontiersin.org
  (13 pages)

Pattern spectrum estimation is described in the following paper:

• Simple Pattern Spectrum Estimation for Fast Pattern Filtering with CoCoNAD
  Christian Borgelt and David Picado-Muiño
  Proc. 13th Int. Symposium on Intelligent Data Analysis, 37-48.
  Springer-Verlag, Berlin/Heidelberg, Germany 2014
  ida_14.pdf (1224 kb) ida_14.ps.gz (1190 kb) (11 pages)