def run(self, max_size=10, variable_size=True, diversify=False, view=config.default_view, **kwargs):
"""
Runs the evolutionary algorithm.
Parameters
----------
max_size : int
Maximum size of a solution, e.g., the maximum number of reactions or genes to knock-out or swap.
variable_size : boolean
If true, the solution size can change meaning that the combination of knockouts can have different sizes up
to max_size. Otherwise it only produces knockout solutions with a fixed number of knockouts.
diversify : bool
It true, the generator will not be allowed to generate repeated candidates in the initial population.
view : cameo.parallel.SequentialView, cameo.parallel.MultiprocessingView
A view for single or multiprocessing.
Returns
-------
TargetOptimizationResult
The result of the optimization.
"""
if kwargs.get('seed', None) is None:
kwargs['seed'] = int(time.time())
self.heuristic_method.observer = self.observers
log_level = simulation_logger.level
simulation_logger.setLevel(logging.CRITICAL)
if diversify:
generator = diversify_function(generators.set_generator)
else:
generator = generators.set_generator
with EvaluatorWrapper(view, self._evaluator) as evaluator:
super(TargetOptimization, self).run(distance_function=set_distance_function,
representation=self.representation,
evaluator=evaluator,
generator=generator,
max_size=max_size,
**kwargs)
simulation_logger.setLevel(log_level)
return TargetOptimizationResult(model=self.model,
heuristic_method=self.heuristic_method,
simulation_method=self.simulation_method,
simulation_kwargs=self._simulation_kwargs,
solutions=self.heuristic_method.archive,
objective_function=self.objective_function,
target_type=self._target_type,
decoder=self._decoder,
evaluator=self._evaluator,
seed=kwargs['seed'],
metadata=self.metadata,
view=view)
def run(self, max_size=10, variable_size=True, diversify=False, view=config.default_view, **kwargs):
"""
Runs the evolutionary algorithm.
Parameters
----------
max_size : int
Maximum size of a solution, e.g., the maximum number of reactions or genes to knock-out or swap.
variable_size : boolean
If true, the solution size can change meaning that the combination of knockouts can have different sizes up
to max_size. Otherwise it only produces knockout solutions with a fixed number of knockouts.
diversify : bool
It true, the generator will not be allowed to generate repeated candidates in the initial population.
view : cameo.parallel.SequentialView, cameo.parallel.MultiprocessingView
A view for single or multiprocessing.
Returns
-------
TargetOptimizationResult
The result of the optimization.
"""
if kwargs.get('seed', None) is None:
kwargs['seed'] = int(time.time())
self.heuristic_method.observer = self.observers
log_level = simulation_logger.level
simulation_logger.setLevel(logging.CRITICAL)
if diversify:
generator = diversify_function(generators.set_generator)
else:
generator = generators.set_generator
with EvaluatorWrapper(view, self._evaluator) as evaluator:
super(TargetOptimization, self).run(distance_function=set_distance_function,
representation=self.representation,
evaluator=evaluator,
generator=generator,
max_size=max_size,
**kwargs)
simulation_logger.setLevel(log_level)
return TargetOptimizationResult(model=self.model,
heuristic_method=self.heuristic_method,
simulation_method=self.simulation_method,
simulation_kwargs=self._simulation_kwargs,
solutions=self.heuristic_method.archive,
objective_function=self.objective_function,
target_type=self._target_type,
decoder=self._decoder,
evaluator=self._evaluator,
seed=kwargs['seed'],
metadata=self.metadata,
view=view)