Provides Context Free Grammar Problems for HeuristicLab

A given BNF grammar is transformed in a SymbolicExpressionGrammar to create GP trees.

Also includes all problem instances from "General Program Synthesis Benchmark Suite" (described in more detail in the technical report: Detailed Problem Descriptions for General Program Synthesis Benchmark Suite)

1. If you want to build HeursiticLab from scratch
   1. Clone HeuristicLab
   2. Build "HeuristicLab.ExtLibs.sln" and then "HeuristicLab 3.3.sln"
   3. Clone HeuristicLab.CFGGP next to HeursticLab
   4. Open and build "HeuristicLab.CFGGP.sln" solution
   5. Binaries will automatically be copied to "HeuristicLab/bin"
2. If you an existing version of HeuristicLab
   1. Clone HeuristicLab.CFGGP
   2. Open and build "HeuristicLab.CFGGP.sln" solution
   3. A folder "HeuristicLab/bin" will be created outside of the repository, which contains the binaries
   4. Copy the binaries in you HeursiticLab version

A new Problem "Context Free Grammar Problem" will be available in the "New Item" Dialog under Problems -> Genetic Programming.

On Windows you can use msbuild. For every other OS use xbuild from Mono, but read documentation How to compile and run HeuristicLab with Mono first.

git clone https://github.com/t-h-e/HeuristicLab.git

msbuild /p:Configuration=Release /p:Platform=x64 HeuristicLab\HeuristicLab.ExtLibs.sln
msbuild /p:Configuration=Release /p:Platform=x64 "HeuristicLab\HeuristicLab 3.3.sln"

git clone https://github.com/t-h-e/HeuristicLab.CFGGP.git

msbuild /p:Configuration=Release /p:Platform=x64 HeuristicLab.CFGGP\HeuristicLab.CFGGP.sln

Executable will be HeuristicLab\bin\HeuristicLab 3.3.exe

Stefan Forstenlechner, David Fagan, Miguel Nicolau and Michael O'Neill Extending Program Synthesis Grammars for Grammar-Guided Genetic Programming PPSN 2018: 15th International Conference on Parallel Problem Solving from Nature, Springer Verlag, 8–12 Sep. 2018.

Stefan Forstenlechner, David Fagan, Miguel Nicolau and Michael O'Neill Towards Effective Semantic Operators for Program Synthesis in Genetic Programming GECCO ’18: Genetic and Evolutionary Computation Conference, ACM, 15–19 July 2018.

Stefan Forstenlechner, David Fagan, Miguel Nicolau and Michael O'Neill Towards Understanding and Refining the General Program Synthesis Benchmark Suite with Genetic Programming CEC 2018: IEEE Congress on Evolutionary Computation, IEEE, 8–13 July 2018.

Stefan Forstenlechner, David Fagan, Miguel Nicolau and Michael O'Neill Semantics-based crossover for program synthesis in genetic programming Artifcial Evolution, Springer Verlag, 25-27 Oct. 2017.

Stefan Forstenlechner, David Fagan, Miguel Nicolau and Michael O'Neill A Grammar Design Pattern for Arbitrary Program Synthesis Problems in Genetic Programming EuroGP 2017: Proceedings of the 20th European Conference on Genetic Programming, LNCS, Vol. 10196, pp. 262-277, Springer Verlag, 19-21 April 2017.

Stefan Forstenlechner, Miguel Nicolau, David Fagan and Michael O'Neill Grammar Design for Derivation Tree Based Genetic Programming Systems EuroGP 2016: Proceedings of the 19th European Conference on Genetic Programming, LNCS, Vol. 9594, pp. 199-214, Springer Verlag, 30 March-1 April 2016.