eckity.genetic_operators.genetic_operator
1import random 2from abc import ABC, abstractmethod 3from overrides import override 4 5from eckity.event_based_operator import Operator 6 7 8class GeneticOperator(Operator, ABC): 9 """ 10 Abstract class for genetic operators. 11 Each operator has a probability of being applied each generation. 12 13 Genetic operators are applied in-place. 14 15 16 Parameters 17 ---------- 18 probability : float, optional 19 probability of being applied each generation, by default 1.0 20 arity : int, optional 21 number of individuals required for the operator, by default 0 22 events : List[str], optional 23 custom events that the operator publishes, by default None 24 """ 25 26 def __init__(self, probability=1.0, arity=0, events=None): 27 super().__init__(events=events, arity=arity) 28 self.probability = probability 29 30 @override 31 def apply_operator(self, payload): 32 """ 33 Apply the genetic operator with a certain probability. 34 The individuals are modified in-place, so it is not mandatory 35 to return them. 36 37 Parameters 38 ---------- 39 payload : List[Individual] 40 Individuals to apply the operator to. 41 42 Returns 43 ------- 44 List[Individual] 45 The individuals after applying the operator. 46 """ 47 if random.random() <= self.probability: 48 # Fitness is irrelevant once the operator is applied 49 for individual in payload: 50 individual.set_fitness_not_evaluated() 51 op_res = self.apply(payload) 52 53 # Add the operator to the applied operators list 54 for ind in op_res: 55 ind.applied_operators.append(type(self).__name__) 56 57 if ind.update_parents: 58 parents = [p.id for p in payload] 59 ind.parents.extend(parents) 60 return op_res 61 return payload 62 63 @abstractmethod 64 def apply(self, individuals): 65 """ 66 Apply the genetic operator to the individuals. 67 This method should be implemented by the subclasses. 68 69 Parameters 70 ---------- 71 individuals : List[Individual] 72 Individuals to apply the operator to. 73 """ 74 pass
9class GeneticOperator(Operator, ABC): 10 """ 11 Abstract class for genetic operators. 12 Each operator has a probability of being applied each generation. 13 14 Genetic operators are applied in-place. 15 16 17 Parameters 18 ---------- 19 probability : float, optional 20 probability of being applied each generation, by default 1.0 21 arity : int, optional 22 number of individuals required for the operator, by default 0 23 events : List[str], optional 24 custom events that the operator publishes, by default None 25 """ 26 27 def __init__(self, probability=1.0, arity=0, events=None): 28 super().__init__(events=events, arity=arity) 29 self.probability = probability 30 31 @override 32 def apply_operator(self, payload): 33 """ 34 Apply the genetic operator with a certain probability. 35 The individuals are modified in-place, so it is not mandatory 36 to return them. 37 38 Parameters 39 ---------- 40 payload : List[Individual] 41 Individuals to apply the operator to. 42 43 Returns 44 ------- 45 List[Individual] 46 The individuals after applying the operator. 47 """ 48 if random.random() <= self.probability: 49 # Fitness is irrelevant once the operator is applied 50 for individual in payload: 51 individual.set_fitness_not_evaluated() 52 op_res = self.apply(payload) 53 54 # Add the operator to the applied operators list 55 for ind in op_res: 56 ind.applied_operators.append(type(self).__name__) 57 58 if ind.update_parents: 59 parents = [p.id for p in payload] 60 ind.parents.extend(parents) 61 return op_res 62 return payload 63 64 @abstractmethod 65 def apply(self, individuals): 66 """ 67 Apply the genetic operator to the individuals. 68 This method should be implemented by the subclasses. 69 70 Parameters 71 ---------- 72 individuals : List[Individual] 73 Individuals to apply the operator to. 74 """ 75 pass
Abstract class for genetic operators. Each operator has a probability of being applied each generation.
Genetic operators are applied in-place.
Parameters
- probability (float, optional): probability of being applied each generation, by default 1.0
- arity (int, optional): number of individuals required for the operator, by default 0
- events (List[str], optional): custom events that the operator publishes, by default None
@override
def
apply_operator(self, payload):
31 @override 32 def apply_operator(self, payload): 33 """ 34 Apply the genetic operator with a certain probability. 35 The individuals are modified in-place, so it is not mandatory 36 to return them. 37 38 Parameters 39 ---------- 40 payload : List[Individual] 41 Individuals to apply the operator to. 42 43 Returns 44 ------- 45 List[Individual] 46 The individuals after applying the operator. 47 """ 48 if random.random() <= self.probability: 49 # Fitness is irrelevant once the operator is applied 50 for individual in payload: 51 individual.set_fitness_not_evaluated() 52 op_res = self.apply(payload) 53 54 # Add the operator to the applied operators list 55 for ind in op_res: 56 ind.applied_operators.append(type(self).__name__) 57 58 if ind.update_parents: 59 parents = [p.id for p in payload] 60 ind.parents.extend(parents) 61 return op_res 62 return payload
Apply the genetic operator with a certain probability. The individuals are modified in-place, so it is not mandatory to return them.
Parameters
- payload (List[Individual]): Individuals to apply the operator to.
Returns
- List[Individual]: The individuals after applying the operator.
@abstractmethod
def
apply(self, individuals):
64 @abstractmethod 65 def apply(self, individuals): 66 """ 67 Apply the genetic operator to the individuals. 68 This method should be implemented by the subclasses. 69 70 Parameters 71 ---------- 72 individuals : List[Individual] 73 Individuals to apply the operator to. 74 """ 75 pass
Apply the genetic operator to the individuals. This method should be implemented by the subclasses.
Parameters
- individuals (List[Individual]): Individuals to apply the operator to.