def GenOffspring(self, pop):
"""
This function generates the offspring by applying crossover, mutation **or** reproduction.
The sum of both probabilities self.cxpb and self.mutpb must be in [0,1]
The reproduction probability is 1 - cxpb - mutpb
The new offspring goes for fitness evaluation
Inputs:
pop (dict): population in dictionary structure
Returns:
offspring (dict): new modified population in dictionary structure
"""
pop_indices = list(range(0, len(pop)))
offspring = defaultdict(list)
for i in range(self.lambda_):
alpha = random.random()
#------------------------------
# Crossover
#------------------------------
if alpha < self.cxpb:
index1, index2 = random.sample(pop_indices, 2)
ind1, ind2, strat1, strat2 = cxES2point(
ind1=list(pop[index1][0]),
ind2=list(pop[index2][0]),
strat1=list(pop[index1][1]),
strat2=list(pop[index2][1]))
offspring[i].append(ind1)
offspring[i].append(strat1)
#print('crossover is done for sample {} between {} and {}'.format(i,index1,index2))
#------------------------------
# Mutation
#------------------------------
elif alpha < (self.cxpb + self.mutpb): # Apply mutation
index = random.choice(pop_indices)
ind, strat = mutES(ind=list(pop[index][0]),
strat=list(pop[index][1]),
lb=self.lbound,
ub=self.ubound,
datatype=self.datatype,
smax=self.smax,
smin=self.smin)
offspring[i].append(ind)
offspring[i].append(strat)
#print('mutation is done for sample {} based on {}'.format(i,index))
#------------------------------
# Reproduction from population
#------------------------------
else:
index = random.choice(pop_indices)
offspring[i].append(pop[index][0])
offspring[i].append(pop[index][1])
#print('reproduction is done for sample {} based on {}'.format(i,index))
return offspring
def GenOffspring(self, pop):
#"""
#
#This function generates the offspring by applying crossover, mutation **or** reproduction.
#The sum of both probabilities self.cxpb and self.mutpb must be in [0,1]
#The reproduction probability is 1 - cxpb - mutpb
#The new offspring goes for fitness evaluation
#Inputs:
# pop (dict): population in dictionary structure
#Returns:
# offspring (dict): new modified population in dictionary structure
#"""
pop_indices = list(range(0, len(pop)))
offspring = defaultdict(list)
for i in range(self.lambda_):
rn = random.random()
#------------------------------
# Crossover
#------------------------------
if rn < self.cxpb:
index1, index2 = random.sample(pop_indices, 2)
if self.cxmode.strip() == 'cx2point':
ind1, ind2, strat1, strat2 = cxES2point(
ind1=list(pop[index1][0]),
ind2=list(pop[index2][0]),
strat1=list(pop[index1][1]),
strat2=list(pop[index2][1]))
elif self.cxmode.strip() == 'blend':
ind1, ind2, strat1, strat2 = cxESBlend(
ind1=list(pop[index1][0]),
ind2=list(pop[index2][0]),
strat1=list(pop[index1][1]),
strat2=list(pop[index2][1]),
alpha=self.alpha)
else:
raise ValueError(
'--error: the cxmode selected (`{}`) is not available in ES, either choose `cx2point` or `blend`'
.format(self.cxmode))
offspring[i].append(ind1)
offspring[i].append(strat1)
#print('crossover is done for sample {} between {} and {}'.format(i,index1,index2))
#------------------------------
# Mutation
#------------------------------
elif rn < self.cxpb + self.mutpb: # Apply mutation
index = random.choice(pop_indices)
ind, strat = self.mutES(ind=list(pop[index][0]),
strat=list(pop[index][1]))
offspring[i].append(ind)
offspring[i].append(strat)
#print('mutation is done for sample {} based on {}'.format(i,index))
#------------------------------
# Reproduction from population
#------------------------------
else:
index = random.choice(pop_indices)
offspring[i].append(pop[index][0])
offspring[i].append(pop[index][1])
#print('reproduction is done for sample {} based on {}'.format(i,index))
return offspring