flamapy-bdd-colosal 2.1.0.dev0

bdd-plugin for the automated analysis of feature models

BDD plugin for Flama (UNED version)

• BDD plugin for Flama (UNED version)
  • Description
  • Requirements and Installation
  • Functionality and usage
    • Load a feature model and create the BDD
    • Save and load a BDD to/from a file
    • Analysis operations
  • Contributing to the BDD plugin

Description

This plugin supports Binary Decision Diagrams (BDDs) representations for feature models.

The plugin is based on flamapy and thus, it follows the same architecture.

The BDD plugin relies on the bbd4va library to synthetize and manipulate BDDs.

Requirements and Installation

• Python 3.9+
• libgmp-dev package
• This plugin depends on the flamapy framework and on the Feature Model plugin.

pip install flamapy-fw flamapy-fm flamapy-bdd

We have tested the plugin on Linux.

Functionality and usage

The executable script test_bdd_metamodel.py serves as an entry point to show the plugin in action.

The following functionality is provided:

Load a feature model and create the BDD

from flamapy.metamodels.fm_metamodel.transformations import UVLReader
from flamapy.metamodels.bdd_metamodel.transformations import FmToBDD

# Load the feature model from UVLReader
feature_model = UVLReader('models/uvl_models/Pizzas.uvl').transform() 

# Create the BDD from the feature model
bdd_model = FmToBDD(feature_model).transform()

Save and load a BDD to/from a file

from flamapy.metamodels.bdd_metamodel.transformations import DDDMPWriter, DDDMPReader
# Save the BDD to a .dddmp file
DDDMPWriter(path='path/to/save/my_bdd.dddmp', source_model=bdd_model).transform()
# Load a BDD from a .dddmp file
bdd_model = DDDMPReader(path='path/to/my_bdd.dddmp').transform()

Analysis operations

• Satisfiable

  Return whether the model is satisfiable (valid):

  from flamapy.metamodels.bdd_metamodel.operations import BDDSatisfiable
  satisfiable = BDDSatisfiable().execute(bdd_model).get_result()
  print(f'Satisfiable? (valid?): {satisfiable}')
  

• Configurations number

  Return the number of configurations:

  from flamapy.metamodels.bdd_metamodel.operations import BDDConfigurationsNumber
  n_configs = BDDConfigurationsNumber().execute(bdd_model).get_result()
  print(f'#Configurations: {n_configs}')
  

• Configurations

  Enumerate the configurations of the model:

  from flamapy.metamodels.bdd_metamodel.operations import BDDConfigurations
  configurations = BDDConfigurations().execute(bdd_model).get_result()
  for i, config in enumerate(configurations, 1):
      print(f'Config {i}: {[feat for feat in config.elements if config.elements[feat]]}')
  

• Sampling

  Return a sample of the given size of uniform random configurations with or without replacement:

  from flamapy.metamodels.bdd_metamodel.operations import BDDSampling
  sampling_op = BDDSampling()
  sampling_op.set_sample_size(5)
  sampling_op.set_with_replacement(False)  # Default False
  sample = sampling_op.execute(bdd_model).get_result()
  for i, config in enumerate(sample, 1):
      print(f'Config {i}: {[feat for feat in config.elements if config.elements[feat]]}')
  

• Product Distribution

  Return the number of products (configurations) having a given number of features:

  from flamapy.metamodels.bdd_metamodel.operations import BDDProductDistribution
  dist = BDDProductDistribution().execute(bdd_model).get_result()
  print(f'Product Distribution: {dist}')
  

• Feature Inclusion Probability

  Return the probability for a feature to be included in a valid configuration:

  from flamapy.metamodels.bdd_metamodel.operations import BDDFeatureInclusionProbability
  prob = BDDFeatureInclusionProbability().execute(bdd_model).get_result()
  for feat in prob.keys():
      print(f'{feat}: {prob[feat]}')
  

• Core features

  Return the core features (those features that are present in all the configurations):

  from flamapy.metamodels.bdd_metamodel.operations import BDDCoreFeatures
  core_features = BDDCoreFeatures().execute(bdd_model).get_result()
  print(f'Core features: {core_features}')
  

• Dead features

  Return the dead features (those features that are not present in any configuration):

  from flamapy.metamodels.bdd_metamodel.operations import BDDDeadFeatures
  dead_features = BDDDeadFeatures().execute(bdd_model).get_result()
  print(f'Dead features: {dead_features}')
  

Contributing to the BDD plugin

To contribute in the development of this plugin:

1. Fork the repository into your GitHub account.
2. Clone the repository: [e-mail]:<<username>>/bdd_metamodel_colosal.git
3. Create a virtual environment: python -m venv env
4. Activate the virtual environment: source env/bin/activate
5. Install the plugin dependencies: pip install flamapy flamapy-fm
6. Install the BDD plugin from the source code: pip install -e bdd_metamodel_colosal

Please try to follow the standards code quality to contribute to this plugin before creating a Pull Request:

• To analyze your Python code and output information about errors, potential problems, convention violations and complexity, pass the prospector with:

  make lint

• To analyze the static type checker for Python and find bugs, pass the Mypy:

  make mypy