Replication package for the paper "On the Use of Multi-valued Decision Diagrams to Count Valid Configurations of Feature Models"

Package structure

• benchmarks:
  • industrial: set of industrial feature models
  • synthetic: set of synthetic feature models
• benchmarks_converter: Java program used to translate the SPLOT benchmarks into the FeatureIDE format. These models are used to reply to RQ1, RQ2, and RQ3, and stored in the benchmarks/synthetic folder
• experiments_results: evaluation scripts and results obtained with our experiments. These data are used to reply to all RQs in our paper
  • images: the folder contains the images produced by evaluation scripts and used in the paper
  • BenchmarksAnalysis_RQ1.ipynb: jupyter notebook used to answer RQ1
  • BenchmarksAnalysis_RQ2.ipynb: jupyter notebook used to answer RQ2
  • BenchmarksAnalysis_RQ3.ipynb: jupyter notebook used to answer RQ3
  • BenchmarksAnalysis_RQ4.ipynb: jupyter notebook used to answer RQ4
  • experiments.csv: it contains the results of the experiments (used in RQ1, RQ2, and RQ3) on synthetic benchmarks
  • FMAnalysis.ipynb: jupyter notebook used to extract the percentage of AND, OR and ALT groups in the industrial benchmarks
• fm_counter_bdd: the folder contains the java program used to compute the number of valid configurations using BDDs
  • dist: the folder contains the executable
    • Counter.jar: executable jar
    • CountWithBDDs.sh: bash script executing Counter.jar over all the industrial benchmarks
  • src: the folder contains the Java source code of the fm_counter_bdd project
• fm_counter_mdd: the folder contains the C++ program used to compute the number of valid configurations using MDDs, i.e., the approach we present in the paper
  • dist: the folder contains the executable
    • FMBuilderExperimenter: executable, compiled for Ubuntu
    • run_experiments.sh: bash script executing FMBuilderExperimenter over all the synthetic benchmarks
  • src: the folder contains the C++ source code of the fm_counter_mdd project