import numpy as np
from typing import List
from desdeo_emo.selection.SelectionBase import SelectionBase
from desdeo_emo.population.Population import Population
[docs]class TournamentSelection(SelectionBase):
"""
Tournament selection operator.
Parameters
----------
pop : Population
The population of individuals
tournament_size : int
Size of the tournament.
"""
def __init__(self, pop, tournament_size):
# initialize
#self.fitness = pop.fitness
#self.pop_size = pop.pop_size
self.tournament_size = tournament_size
# TODO: add the opt fitness, which is the custom set of fitness values e.g self.local_fitnesss
[docs] def do(self, pop, fitness) -> List[int]:
""" Performs tournament selections and returns the parents.
Parameters
----------
pop : Population
The current population.
Returns
-------
List[int]
List of indices of the selected individuals
"""
self.fitness = fitness
self.pop_size = self.fitness.shape[0]
parents = []
for i in range(int(self.pop_size)):
parents.append(
np.asarray(
self._tour_select(),
self._tour_select(),
))
return parents
[docs] def _tour_select(self):
"""Tournament selection. Choose number of individuals to participate
and select the one with the best fitness.
Parameters
----------
fitness : array_like
An array of each individual's fitness.
tournament_size : int
Number of participants in the tournament.
Returns
-------
int
The index of the best individual.
"""
# make next 4 lines prettier and check if PBEA article uses Deterministic tournament selection
fitness = self.fitness[:,0]
aspirants = np.random.choice(len(fitness)-1, self.tournament_size, replace=False)
chosen = []
for ind in aspirants:
chosen.append([ind, fitness[ind]])
chosen.sort(key=lambda x: x[1])
return chosen[0][0]
