rocreguant

#!/bin/bash
#
# https://github.com/Nyr/openvpn-install
#
# Copyright (c) 2013 Nyr. Released under the MIT License.


# Detect Debian users running the script with "sh" instead of bash
if readlink /proc/$$/exe | grep -q "dash"; then
	echo 'This installer needs to be run with "bash", not "sh".'

rocreguant

import numpy as np

def roulette_wheel_selection(population):
  
    # Computes the totallity of the population fitness
    population_fitness = sum([chromosome.fitness for chromosome in population])
    
    # Computes for each chromosome the probability 
    chromosome_probabilities = [chromosome.fitness/population_fitness for chromosome in population]
    

rocreguant

import numpy as np

def roulette_wheel_selection(population):
  
    # Computes the totallity of the population fitness
    population_fitness = sum([chromosome.fitness for chromosome in population])
    
    # Computes for each chromosome the probability 
    chromosome_probabilities = [chromosome.fitness/population_fitness for chromosome in population]
    

rocreguant

best_solution = [-1,-np.inf,np.array([])]
for i in range(10000):
    if i%100==0: print(i, fitnes_list.min(), fitnes_list.mean(), datetime.now().strftime("%d/%m/%y %H:%M"))
    fitnes_list = get_all_fitnes(mutated_pop,cities_dict)
    
    #Saving the best solution
    if fitnes_list.max() > best_solution[1]:
        best_solution[0] = i
        best_solution[1] = fitnes_list.max()
        best_solution[2] = np.array(mutated_pop)[fitnes_list.max() == fitnes_list]

rocreguant

def mutate_offspring(offspring, max_weight, object_list, objects_dict):
    for mutation_number in range(int(len(offspring)*mutation_rate)):
    
        a = np.random.randint(0,len(object_list))
        b = np.random.randint(0,len(offspring))
        
        offspring.insert(b, object_list[a])
        
    offspring = fit_in_knapsack(offspring, max_weight, objects_dict)

rocreguant

def mate_progenitors(prog_a, prog_b, max_weight, objects_dict):
    offspring = []

    for i in zip(prog_a, prog_b):
        offspring.extend(i)
        offspring = list(dict.fromkeys(offspring)) #Removing duplicates
        offspring = fit_in_knapsack(offspring, max_weight, objects_dict)

    return offspring
            

rocreguant

def progenitor_selection(population_set,fitnes_list):
    total_fit = fitnes_list.sum()
    prob_list = fitnes_list/total_fit
    
    #Notice there is the chance that a progenitor. mates with oneself
    progenitor_list_a = np.random.choice(list(range(len(population_set))), len(population_set),p=prob_list, replace=True)
    progenitor_list_b = np.random.choice(list(range(len(population_set))), len(population_set),p=prob_list, replace=True)
    
    progenitor_list_a = population_set[progenitor_list_a]
    progenitor_list_b = population_set[progenitor_list_b]

rocreguant

def get_all_fitnes(population_set, objects_dict):
    fitnes_list = np.zeros(n_population)

    #Looping over all solutions computing the fitness for each solution
    for i in  range(n_population):
        _, fitnes_list[i] = get_current_weight_value(population_set[i], objects_dict)

    return fitnes_list

rocreguant

# First step: Create the first population set
def fit_in_knapsack(objects_list, max_weight, objects_dict):
    r = []
    for o in objects_list:
        r.append(o)
        weight, value = get_current_weight_value(r, objects_dict)
        if weight > max_weight:
            r.pop()
            return r
    return r

rocreguant

# Everything put together
best_solution = [-1,np.inf,np.array([])]
for i in range(100000):
    if i%100==0: print(i, fitnes_list.min(), fitnes_list.mean(), datetime.now().strftime("%d/%m/%y %H:%M"))
    fitnes_list = get_all_fitnes(mutated_pop,cities_dict)
    
    #Saving the best solution
    if fitnes_list.min() < best_solution[1]:
        best_solution[0] = i
        best_solution[1] = fitnes_list.min()