opensbt.experiment package

Submodules

opensbt.experiment.experiment module

class opensbt.experiment.experiment.Experiment(name: str, problem: SimulationProblem, algorithm: str, search_configuration: SearchConfiguration)

Bases: object

The experiment class holds all information to execute a testing experiment and allows to maintain internally different experiment configurations to seperate experiments from each other.

__init__(name: str, problem: SimulationProblem, algorithm: str, search_configuration: SearchConfiguration)

Create an Experiment.

Parameters:

• name (str) – Name of the experiment to identify it.

• problem (Problem) – The corresponding testing problem.

• algorithm (str) – The algorithm string tag (enum) to use for testing.

• search_configuration (SearchConfiguration) – The search configuration.

opensbt.experiment.experiment_store module

class opensbt.experiment.experiment_store.DefaultExperiments(*args, **kwds)

Bases: Singleton

This class allows to store and load experiments globally in OpenSBT.

get_store()

Outputs the store.

Returns:

Return the store.

Return type:

Dict

init()

Initializes the store.

load(experiment_name: str) → Experiment

Loads an experiment based on the experiment name from the store.

Parameters:

experiment_name (str) – The experiment name.

Returns:

Returns the experiment.

Return type:

Experiment

register(exp: Experiment)

Registers an experiment in the store.

Parameters:

exp (Experiment) – The experiment to be registered.

Returns:

Output 1 if registration failed, otherwise 0.

Return type:

int

class opensbt.experiment.experiment_store.Singleton(*args, **kwds)

Bases: object

init(*args, **kwds)

opensbt.experiment.search_configuration module

class opensbt.experiment.search_configuration.DefaultSearchConfiguration

Bases: SearchConfiguration

This class holds all configuration parameter initialized with default values

custom_params = {}

eta_crossover = 20

eta_mutation = 15

inner_num_gen = 4

max_tree_iterations = 4

maximal_execution_time = None

n_func_evals_lim = 500

n_generations = 5

nadir = None

num_offsprings = None

operators = {'cx': None, 'dup': None, 'init': None, 'mut': None}

population_size = 20

prob_crossover = 0.7

prob_mutation = None

ref_point_hv = None

seed = None

class opensbt.experiment.search_configuration.SearchConfiguration

Bases: object

This class holds all configuration parameter related to opimization algorithms

__init__() → None

custom_params = {}

eta_crossover = None

eta_mutation = None

ideal = None

inner_num_gen = None

max_tree_iterations = None

maximal_execution_time = None

n_func_evals_lim = None

n_generations = None

nadir = None

num_offsprings = None

operators = {'cx': None, 'dup': None, 'init': None, 'mut': None}

population_size = None

prob_crossover = None

prob_mutation = None

ref_point_hv = None

seed = None

Module contents