Skip to content

thomaskeck/NeurobayesCppInterface

 
 

Repository files navigation

NeuroBayes C++ Interface

Interface to NeuroBayes(R) for academic purpose

The NeuroBayes package is available under the NeuroBayes license from Blue Yonder GmbH. Please contact neurobayes@blue-yonder.com for general information and for information about the license.

This repository provides the C++ Interface of the NeuroBayes package. You can only build this project if you obtained the NeuroBayes licence and NeuroBayes Core library from Blue Yonder GmbH.

The NeuroBayes C++ Interface is usually already shipped with the NeuroBayes package. The NeuroBayes Core library does only depenend on a very small set of system libraries like libc and libgfortran, on the other hand the NeuroBayes C++ Interface depends on libstdc++. It may be necessary to recompile the C++ Interface in the future to support new platforms, hence Blue Yonder GmbH decided to release the source code of the C++ Interface under the MIT licence for the convinience of the NeuroBayes user community (in particular for high energy physics experiments).

Build process and installation

  • Extract the NeuroBayes Core shared library named libNeuroBayesCore_shared.so from the package you obtained from Blue Yonder GmbH and copy it into the previously empty directory core.
  • cd build/
  • cmake ../
  • make

Building rpm, deb or tgz packages

Optionally you can build rpm, deb and tgz packages using

  • make package Additional third-party libraries are required to do so.

Maxnode

The NeuroBayes Core library is optimized for a maximum number of nodes. The maximum number of nodes is printed by the NeuroBayes Teacher at the start of the program e.g. execute the minimaltest executable and search for NB_MAXNODE. The default number is 100. If your library is optimized for a different number of nodes, you have to provide the correct number using an environment variable named MAXNODE_PREPRO. during the build process e.g. MAXNODE_PREPRO=200 cmake ../

Tests

To verify the correctness of your executable, you can execute the following test suits:

  • cd build
  • ./minimaltest
  • ./interfacetest

The minimaltest provides also a minimal example to get you started.

About

Interface to NeuroBayes(R) for academic purpose

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 86.1%
  • C 11.5%
  • CMake 2.4%