def generateChild(isValidPop, indexParent, incomp, travels, links, nbBus,
populationParent, adnCroisementCount_min,
adnCroisementCount_max):
childGenes = []
if not isValidPop:
croisementIndex = randint(
1,
len(populationParent[indexParent][0].adn) - 2)
childGenes = populationParent[indexParent][
0].adn[:croisementIndex] + populationParent[indexParent][1].adn[-(
len(populationParent[indexParent][1].adn) - croisementIndex):]
else:
indexParentSelected = randint(0,
len(populationParent[indexParent]) - 1)
indexParentSelected2 = randint(
0,
len(
set(populationParent[indexParent]) -
set([int(indexParentSelected)])) - 1)
childGenes = list(
populationParent[indexParent][indexParentSelected].adn)
for i in range(randint(adnCroisementCount_min,
adnCroisementCount_max)):
croisementIndex = randint(
0,
len(populationParent[indexParent][0].adn) - 1)
childGenes[croisementIndex] = populationParent[indexParent][
indexParentSelected2].adn[croisementIndex]
child = Individu(childGenes)
if isValidPop:
child.computeScoreNEW(incomp, travels, links, nbBus)
return child
def generateChild(
isValidPop,
indexParent,
incomp,
travels,
links,
nbBus,
populationParent,
adnCroisementCount_min,
adnCroisementCount_max,
):
childGenes = []
if not isValidPop:
croisementIndex = randint(1, len(populationParent[indexParent][0].adn) - 2)
childGenes = (
populationParent[indexParent][0].adn[:croisementIndex]
+ populationParent[indexParent][1].adn[-(len(populationParent[indexParent][1].adn) - croisementIndex) :]
)
else:
indexParentSelected = randint(0, len(populationParent[indexParent]) - 1)
indexParentSelected2 = randint(0, len(set(populationParent[indexParent]) - set([int(indexParentSelected)])) - 1)
childGenes = list(populationParent[indexParent][indexParentSelected].adn)
for i in range(randint(adnCroisementCount_min, adnCroisementCount_max)):
croisementIndex = randint(0, len(populationParent[indexParent][0].adn) - 1)
childGenes[croisementIndex] = populationParent[indexParent][indexParentSelected2].adn[croisementIndex]
child = Individu(childGenes)
if isValidPop:
child.computeScoreNEW(incomp, travels, links, nbBus)
return child
def __init__(self, nombre_individus, mot_a_trouver):
self._phrase = mot_a_trouver
self._taille_pop = nombre_individus
for _ in range(self._taille_pop):
individu = Individu.Individu()
individu.generate_phrase(len(mot_a_trouver))
individu.calcul_fitness(mot_a_trouver)
self._individus.append(individu)
def createIndividu(travels):
travelsTMP = []
for i in range(len(travels)):
travelsTMP.append(i)
gene = []
for i in range(len(travels)):
index = randint(0, len(travelsTMP) - 1)
gene.append(travelsTMP[index])
del travelsTMP[index]
return Individu(gene)
def createIndividu2(travels, nbBus):
idBus = []
gene = []
for i in range(len(travels)):
idBus.append(i)
for i in range(len(travels)):
val = randint(0, len(idBus) - 1)
gene.append(idBus[val])
del idBus[val]
return Individu(gene)
def generateChildren(populationParent,adnCroisementCount):
children = []
for i in range(len(populationParent)):
croisementIndex = randint(1,len(populationParent[i][0].adn.genes)-1)
childGenes = []
for j in range(len(populationParent[i][0].adn.genes)):
if j < croisementIndex :
childGenes.append(populationParent[i][0].adn.genes[j])
else :
childGenes.append(populationParent[i][1].adn.genes[j])
# print str(i) + '>>>' + str(childGenes) + '(' + str(croisementIndex) + ')'
children.append(Individu(ADN(childGenes)))
return children
def fillGroup(self):
for i in range(self.SIZE):
age = random.randrange(0, self.MAXAGE) + 1
school = random.randrange(0, self.N_SCHOOL)
work = random.randrange(0, self.N_ENTERPRISE)
family = random.randrange(0, self.SIZE / self.FAMILYSIZE)
friendgroup = random.randrange(0, self.N_FRIENDGROUP)
self.group.append(Individu(age, school, work, family, friendgroup))
self.families[family].add(self.group[i])
self.friends[friendgroup].add(self.group[i])
if self.group[i].isStudent():
self.schools[school].add(self.group[i])
elif self.group[i].isWorking():
self.enterprises[work].add(self.group[i])
def __init__(self, nb_individus, graph_type, ind_size):
CC_ref = 10000
self.nb_individus = nb_individus
self.ind_size = ind_size
self.pop = [] # Contiendra tous les individus
for i in range(nb_individus):
self.pop.append(Individu(graph_type, ind_size))
#liste des fitness moyennes incrementée chaque pas de temps
self.WMOY = []
self.WMOY_CC = []
self.WMOY_DI = []
self.WMOY_DD = []
self.Maj_fitness()
def crossover_pop2(self):
mot_bebe = ""
cross = []
for _ in range(len(self._individus)):
p1, p2 = random.sample(self._selection_pop, 2)
p1 = p1.get_phrase()
p2 = p2.get_phrase()
for i in range(len(p1)):
if random.randrange(0, 2) == 1:
mot_bebe += p1[i]
else:
mot_bebe += p2[i]
ind = Individu.Individu()
ind.set_phrase(mot_bebe)
ind.calcul_fitness(self._phrase)
cross.append(ind)
mot_bebe = ""
self._individus = cross
def crossover_pop1(self):
new_pop = []
for _ in range(self._taille_pop):
# Génération de deux parents choisis dans la population sélectionnée
parent_1, parent_2 = random.sample(self._selection_pop, 2)
parent_1 = parent_1.get_phrase()
parent_2 = parent_2.get_phrase()
# Aléatoire sur taille du parent moins 2
gene_parent = random.randrange(1, len(parent_1) - 1)
phrase_enfant = parent_1[0:gene_parent] + parent_2[gene_parent:]
# Création d'un enfant de deux individus auquel on attribue la phrase généré
enfant_individu = Individu.Individu()
enfant_individu.set_phrase(phrase_enfant)
enfant_individu.calcul_fitness(self._phrase)
# Ajout de cet enfant dans la nouvelle population
new_pop.append(enfant_individu)
self._individus = new_pop
def crossing_over(
self, G1, G2
): #Prend 2 graphes parents en argument en retourne un graph enfant
n1 = rn.randint(
self.ind_size / 4, 3 * self.ind_size /
4) #Nombre de noeuds provenant du graphe 1 choisis aleatoirement
Noeuds1 = rn.sample(list(G1.G.nodes),
n1) #Liste des noeuds de n1 choisis
Noeuds2 = []
for n in range(self.ind_size
): #Remplissage de la liste des noeuds provenant de G2
if n not in Noeuds1:
Noeuds2.append(n)
#print "noeuds du g1 :", Noeuds1
#print "noeuds du g2 :", Noeuds2
Edges = [] #Liste des edges des noeuds de G1
for noeud in Noeuds1:
for e in list(G1.G.edges(noeud)):
Edges.append(e)
Edges2 = [] #Liste des edges des noeuds de G2
for noeud in Noeuds2:
for e in list(G2.G.edges(noeud)):
Edges2.append(e)
G3 = Individu("NONE", self.ind_size) #Initialisation du graph enfant
G3.G.add_nodes_from(
G1.G) #Possede le meme nombre de noeuds que les graphes parents
G3.G.add_edges_from(Edges) #Prend les edges de G1
G3.G.add_edges_from(Edges2) # Prend les edges de G2
#Met a jour les attributs et la fitness du graph
G3.maj_attributs()
G3.maj_fitness()
#print "G3 : ",G3.CC, G3.DI, G3.DD
return G3 #Retourne le graph enfant
def createIndividu(): gene = [] for i in range(16): gene.append(randint(0,1)) return Individu(ADN(gene))
import time as tm import random as rd import os import pickle as pk import Individu as ind from tkinter import * jean = ind.Individu(20, 5, 6, 80, True) marie = ind.Individu(20, 5, 6, 120, False) bebe = ind.enfant(jean, marie)
def createIndividu(travels, nbBus):
gene = []
for i in range(len(travels)):
gene.append(randint(0, nbBus - 1))
return Individu(gene)