Source code for pymoo.core.indicator

import abc

from pymoo.util.normalization import PreNormalization




class Indicator(PreNormalization):

    def __init__(self, **kwargs):
        super().__init__(**kwargs)

        # what should an indicator return if no solutions are provided is defined here
        self.default_if_empty = 0.0

    def __call__(self, F, *args, **kwargs):
        return self.do(F, *args, **kwargs)

    def do(self, F, *args, **kwargs):

        # if it is a 1d array
        if F.ndim == 1:
            F = F[None, :]

        # if no points have been provided just return the default
        if len(F) == 0:
            return self.default_if_empty

        # do the normalization - will only be done if zero_to_one is enabled
        F = self.normalization.forward(F)

        return self._do(F, *args, **kwargs)

    @abc.abstractmethod
    def _do(self, F, *args, **kwargs):
        return