GAopt

A python package for genetic algorithm with parallel processing implemented

The package has the ability for parallel processing and resuming.

example of implementation

import numpy as np
from GAopt import GA
def objective(X):
    X1 = X[0]
    X2 = X[1]
    X3 = X[2]
    X4 = X[3]
    return (X1+X2)/(X3+X4+0.5)

varbound=np.array([[1,3],[1,4],[0.5,1.5],[2,20],])
vartype=np.array([['int'],['int'],['real'],['int'],])

parameters = {'max_num_iteration': None,
              'population_size': 400,
              'mutation_probability':0.1,
              'elit_ratio': 0.1,
              'crossover_probability': 0.5,
              'parents_portion': 0.3,
              'crossover_type':'uniform',
              'max_iteration_without_improv':None,
              'Number_of_processes':'max',
              'Population_file_path': "pop.csv"}

Genetic = GA(objective,4,
             variable_type_mixed=vartype, 
             variable_boundaries=varbound, 
             function_timeout=5000, 
             algorithm_parameters=parameters)

# this line is to run the code for the first time
Genetic.run()

# this line is to resume an already existing run
Genetic.resume("old_pop.csv")