Coverage for moptipy / algorithms / so / fitnesses / direct.py: 100%

1""" 

2Direct fitness assignment uses the objective values directly as fitness. 

3 

4Together with fitness proportionate selection (:class:`~moptipy.algorithms.\ 

5modules.selections.fitness_proportionate_sus.FitnessProportionateSUS`), it 

6turns an EA (:class:`~moptipy.algorithms.so.general_ea.GeneralEA`) to a 

7minimization variant of the traditional Genetic Algorithms. 

8 

91. John Henry Holland. *Adaptation in Natural and Artificial Systems: An 

10 Introductory Analysis with Applications to Biology, Control, and Artificial 

11 Intelligence.* Ann Arbor, MI, USA: University of Michigan Press. 1975. 

12 ISBN: 0-472-08460-7 

132. David Edward Goldberg. *Genetic Algorithms in Search, Optimization, and 

14 Machine Learning.* Boston, MA, USA: Addison-Wesley Longman Publishing Co., 

15 Inc. 1989. ISBN: 0-201-15767-5 

16""" 

17 

18 

19from numpy.random import Generator 

20 

21from moptipy.algorithms.so.fitness import Fitness, FRecord 

22 

23 

24# start book 

25class Direct(Fitness): 

26 """ 

27 Objective values are used as fitness directly. 

28 

29 This is the most primitive and basic fitness assignment strategy. 

30 It does not give preference to new solutions over old solutions if both 

31 have the same objective value. In this, it is different from 

32 :class:`~moptipy.algorithms.so.fitnesses.rank_and_iteration\ 

33.RankAndIteration`, which is the default fitness assignment strategy. 

34 It is therefore also not compatible to our basic (mu+lambda) Evolutionary 

35 Algorithm implementation (:class:`~moptipy.algorithms.so.ea.EA`). 

36 """ 

37 

38 def assign_fitness(self, p: list[FRecord], random: Generator) -> None: 

39 """ 

40 Assign the objective value as fitness. 

41 

42 :param p: the list of records 

43 :param random: ignored 

44 """ 

45 for rec in p: # iterate over all records in the population 

46 rec.fitness = rec.f # store objective value f as fitness 

47# end book 

48 

49 def __str__(self): 

50 """ 

51 Get the name of the direct fitness assignment. 

52 

53 :return: the name of this fitness assignment strategy 

54 :retval "direct": always 

55 """ 

56 return "direct"