observer.py

import logging
import os
from pathlib import Path
from typing import List, TypeVar

import numpy as np
from tqdm import tqdm

from jmetal.core.observer import Observer
from jmetal.core.problem import DynamicProblem
from jmetal.core.quality_indicator import InvertedGenerationalDistance
from jmetal.lab.visualization import Plot, StreamingPlot
from jmetal.util.solution import print_function_values_to_file

S = TypeVar("S")

LOGGER = logging.getLogger("jmetal")

"""
.. module:: observer
   :platform: Unix, Windows
   :synopsis: Implementation of algorithm's observers.

.. moduleauthor:: Antonio J. Nebro <antonio@lcc.uma.es>
"""


class ProgressBarObserver(Observer):
    def __init__(self, max: int) -> None:
        """Show a smart progress meter with the number of evaluations and computing time.

        :param max: Number of expected iterations.
        """
        self.progress_bar = None
        self.progress = 0
        self._max = max

    def update(self, *args, **kwargs):
        if not self.progress_bar:
            self.progress_bar = tqdm(total=self._max, ascii=True, desc="Progress")

        evaluations = kwargs["EVALUATIONS"]

        self.progress_bar.update(evaluations - self.progress)
        self.progress = evaluations

        if self.progress >= self._max:
            self.progress_bar.close()


class BasicObserver(Observer):
    def __init__(self, frequency: int = 1) -> None:
        """Show the number of evaluations, the best fitness and the computing time.
        :param frequency: Display frequency."""

        self.display_frequency = frequency

    def update(self, *args, **kwargs):
        computing_time = kwargs["COMPUTING_TIME"]
        evaluations = kwargs["EVALUATIONS"]
        solutions = kwargs["SOLUTIONS"]

        if (evaluations % self.display_frequency) == 0 and solutions:
            if type(solutions) == list:
                fitness = solutions[0].objectives
            else:
                fitness = solutions.objectives

            LOGGER.info(
                "Evaluations: {} \n Best fitness: {} \n Computing time: {}".format(evaluations, fitness, computing_time)
            )


class PrintObjectivesObserver(Observer):
    def __init__(self, frequency: int = 1) -> None:
        """Show the number of evaluations, best fitness and computing time.

        :param frequency: Display frequency."""
        self.display_frequency = frequency

    def update(self, *args, **kwargs):
        evaluations = kwargs["EVALUATIONS"]
        solutions = kwargs["SOLUTIONS"]

        if (evaluations % self.display_frequency) == 0 and solutions:
            if type(solutions) == list:
                fitness = solutions[0].objectives
            else:
                fitness = solutions.objectives

            LOGGER.info("Evaluations: {}. fitness: {}".format(evaluations, fitness))


class WriteFrontToFileObserver(Observer):
    def __init__(self, output_directory: str) -> None:
        """Write function values of the front into files.

        :param output_directory: Output directory. Each front will be saved on a file `FUN.x`."""
        self.counter = 0
        self.directory = output_directory

        if Path(self.directory).is_dir():
            LOGGER.warning("Directory {} exists. Removing contents.".format(self.directory))
            for file in os.listdir(self.directory):
                os.remove("{0}/{1}".format(self.directory, file))
        else:
            LOGGER.warning("Directory {} does not exist. Creating it.".format(self.directory))
            Path(self.directory).mkdir(parents=True)

    def update(self, *args, **kwargs):
        problem = kwargs["PROBLEM"]
        solutions = kwargs["SOLUTIONS"]

        if solutions:
            if isinstance(problem, DynamicProblem):
                termination_criterion_is_met = kwargs.get("TERMINATION_CRITERIA_IS_MET", None)

                if termination_criterion_is_met:
                    print_function_values_to_file(solutions, "{}/FUN.{}".format(self.directory, self.counter))
                    self.counter += 1
            else:
                print_function_values_to_file(solutions, "{}/FUN.{}".format(self.directory, self.counter))
                self.counter += 1


class PlotFrontToFileObserver(Observer):
    def __init__(self, output_directory: str, step: int = 100, **kwargs) -> None:
        """Plot and save Pareto front approximations into files.

        :param output_directory: Output directory.
        """
        self.directory = output_directory
        self.plot_front = Plot(title="Pareto front approximation", **kwargs)
        self.last_front = []
        self.fronts = []
        self.counter = 0
        self.step = step

        if Path(self.directory).is_dir():
            LOGGER.warning("Directory {} exists. Removing contents.".format(self.directory))
            for file in os.listdir(self.directory):
                os.remove("{0}/{1}".format(self.directory, file))
        else:
            LOGGER.warning("Directory {} does not exist. Creating it.".format(self.directory))
            Path(self.directory).mkdir(parents=True)

    def update(self, *args, **kwargs):
        problem = kwargs["PROBLEM"]
        solutions = kwargs["SOLUTIONS"]
        evaluations = kwargs["EVALUATIONS"]

        if solutions:
            if (evaluations % self.step) == 0:
                if isinstance(problem, DynamicProblem):
                    termination_criterion_is_met = kwargs.get("TERMINATION_CRITERIA_IS_MET", None)

                    if termination_criterion_is_met:
                        if self.counter > 0:
                            igd = InvertedGenerationalDistance(np.array([s.objectives for s in self.last_front]))
                            igd_value = igd.compute(np.array([s.objectives for s in solutions]))
                        else:
                            igd_value = 1

                        if igd_value > 0.005:
                            self.fronts += solutions
                            self.plot_front.plot(
                                [self.fronts],
                                label=problem.get_name(),
                                filename=f"{self.directory}/front-{evaluations}",
                            )
                        self.counter += 1
                        self.last_front = solutions
                else:
                    self.plot_front.plot(
                        [solutions],
                        label=f"{evaluations} evaluations",
                        filename=f"{self.directory}/front-{evaluations}",
                    )
                    self.counter += 1


class VisualizerObserver(Observer):
    def __init__(
        self, reference_front: List[S] = None, reference_point: list = None, display_frequency: int = 1
    ) -> None:
        self.figure = None
        self.display_frequency = display_frequency

        self.reference_point = reference_point
        self.reference_front = reference_front

    def update(self, *args, **kwargs):
        evaluations = kwargs["EVALUATIONS"]
        solutions = kwargs["SOLUTIONS"]

        if solutions:
            if self.figure is None:
                self.figure = StreamingPlot(reference_point=self.reference_point, reference_front=self.reference_front)
                self.figure.plot(solutions)

            if (evaluations % self.display_frequency) == 0:
                # check if reference point has changed
                reference_point = kwargs.get("REFERENCE_POINT", None)

                if reference_point:
                    self.reference_point = reference_point
                    self.figure.update(solutions, reference_point)
                else:
                    self.figure.update(solutions)

                self.figure.ax.set_title("Eval: {}".format(evaluations), fontsize=13)