1 | from abc import ABC |
---|

2 | |
---|

3 | import numpy as np |
---|

4 | |
---|

5 | from evolalg.base.frams_step import FramsStep |
---|

6 | from evolalg.dissimilarity.dissimilarity import Dissimilarity |
---|

7 | |
---|

8 | |
---|

9 | class FramsDissimilarity(FramsStep): |
---|

10 | |
---|

11 | def __init__(self, frams_lib, reduction="mean", output_field="dissim", *args, **kwargs): |
---|

12 | super(FramsDissimilarity, self).__init__(frams_lib, *args, **kwargs) |
---|

13 | |
---|

14 | self.output_field = output_field |
---|

15 | self.fn_reduce = None |
---|

16 | if reduction == "mean": |
---|

17 | self.fn_reduce = np.mean |
---|

18 | elif reduction == "max": |
---|

19 | self.fn_reduce = np.max |
---|

20 | elif reduction == "min": |
---|

21 | self.fn_reduce = np.min |
---|

22 | elif reduction == "sum": |
---|

23 | self.fn_reduce = np.sum |
---|

24 | elif reduction == "none" or reduction is None: |
---|

25 | self.fn_reduce = None |
---|

26 | else: |
---|

27 | raise ValueError("Unknown reduction type. Supported: mean, max, min, sum, none") |
---|

28 | |
---|

29 | def reduce(self, dissim_matrix): |
---|

30 | if self.fn_reduce is None: |
---|

31 | return dissim_matrix |
---|

32 | return self.fn_reduce(dissim_matrix, axis=1) |
---|

33 | |
---|

34 | def call(self, population): |
---|

35 | if len(population) == 0: |
---|

36 | return [] |
---|

37 | dissim_matrix = self.frams.dissimilarity([_.genotype for _ in population]) |
---|

38 | dissim = self.reduce(dissim_matrix) |
---|

39 | for d,ind in zip(dissim, population): |
---|

40 | setattr(ind, self.output_field, d) |
---|

41 | return population |
---|