Source code for macop.callbacks.BasicCheckpoint

"""Basic Checkpoint class implementation
"""

# main imports
import os
import logging
import numpy as np

# module imports
from .Callback import Callback
from ..utils.color import macop_text, macop_line


[docs]class BasicCheckpoint(Callback): """ BasicCheckpoint is used for loading previous computations and start again after loading checkpoint Attributes: algo: {Algorithm} -- main algorithm instance reference every: {int} -- checkpoint frequency used (based on number of evaluations) filepath: {str} -- file path where checkpoints will be saved """
[docs] def run(self): """ Check if necessary to do backup based on `every` variable """ # get current best solution solution = self.algo.bestSolution currentEvaluation = self.algo.getGlobalEvaluation() # backup if necessary if currentEvaluation % self.every == 0: logging.info("Checkpoint is done into " + self.filepath) solutionData = "" solutionSize = len(solution.data) for index, val in enumerate(solution.data): solutionData += str(val) if index < solutionSize - 1: solutionData += ' ' line = str(currentEvaluation) + ';' + solutionData + ';' + str( solution.fitness()) + ';\n' # check if file exists if not os.path.exists(self.filepath): with open(self.filepath, 'w') as f: f.write(line) else: with open(self.filepath, 'a') as f: f.write(line)
[docs] def load(self): """ Load last backup line of solution and set algorithm state (best solution and evaluations) at this backup """ if os.path.exists(self.filepath): logging.info('Load best solution from last checkpoint') with open(self.filepath) as f: # get last line and read data lastline = f.readlines()[-1] data = lastline.split(';') # get evaluation information globalEvaluation = int(data[0]) if self.algo.parent is not None: self.algo.parent.numberOfEvaluations = globalEvaluation else: self.algo.numberOfEvaluations = globalEvaluation # get best solution data information solutionData = list(map(int, data[1].split(' '))) self.algo.bestSolution.data = np.array(solutionData) self.algo.bestSolution.score = float(data[2]) print(macop_line()) print( macop_text('Checkpoint found from `{}` file.'.format( self.filepath))) print( macop_text('Restart algorithm from evaluation {}.'.format( self.algo.numberOfEvaluations))) else: print( macop_text( 'No backup found... Start running algorithm from evaluation 0.' )) logging.info( "Can't load backup... Backup filepath not valid in Checkpoint") print(macop_line())