def __init__(self, problem, num_of_generations, num_of_individuals):
self.utils = NSGA2Utils(problem, num_of_individuals)
self.population = None
self.num_of_generations = num_of_generations
self.on_generation_finished = []
self.num_of_individuals = num_of_individuals
def __init__(self, problem, num_of_generations, num_of_individuals):
print("Initializing Evolution Process Elements")
self.utils = NSGA2Utils(problem, num_of_individuals)
self.population = None
self.num_of_generations = num_of_generations #"""200 How much the Genetic Algo goes one for"""
self.on_generation_finished = []
self.num_of_individuals = num_of_individuals #"""200 How many Chromosomes"""
def __init__(self,
problem,
num_of_generations=1000,
num_of_individuals=100,
num_of_tour_particips=2,
tournament_prob=0.9,
p_c=0.6,
p_m=0.3,
p_risked=0.5,
full=True):
"""
Attributes
----------
utils : NSGA2Utils
class of utile functions
population : Population
population class
num_of_generations : int
number of generations
num_of_individuals : int
population size
problem : Problem
problem definition class
num_of_tour_particips : int
number of indivisuals per selection tournament
tournament_prob : float
paramters in binary selection tournament (not used for now)
p_c : float
crossover probability
p_m : float
mutaion probability
p_risked : float
probability of doing the crossover on the risky assets
full : boolean
True: cross over on the whole portfolio , False : cross over on a part of the portfolio
"""
self.utils = NSGA2Utils(problem, num_of_individuals,
num_of_tour_particips, tournament_prob, p_c,
p_m, p_risked, full)
self.population = None
self.num_of_generations = num_of_generations
self.on_generation_finished = []
self.num_of_individuals = num_of_individuals
self.problem = problem
def __init__(self, problem, evolutions_df, dataset_name, protected_variable,num_of_generations=5, num_of_individuals=10, num_of_tour_particips=2, tournament_prob=0.9, crossover_param=2, mutation_param=5, mutation_prob=0.3, beta_method="uniform"):
self.utils = NSGA2Utils(problem, num_of_individuals, num_of_tour_particips, tournament_prob, crossover_param, mutation_param, mutation_prob, beta_method)
self.population = None
self.evolutions_df = evolutions_df
self.dataset_name = dataset_name
self.protected_variable = protected_variable
self.num_of_generations = num_of_generations
self.on_generation_finished = []
self.num_of_individuals = num_of_individuals
self.mutation_prob = mutation_prob
self.beta_method = beta_method
def __init__(self, problem, vardim, iteration, num_of_generations,
num_of_individuals, mutation_prob, num_of_genes_to_mutate,
num_of_tour_particips):
self.utils = NSGA2Utils(problem, vardim, num_of_generations,
num_of_individuals, mutation_prob,
num_of_genes_to_mutate, num_of_tour_particips)
self.population = None
self.vardim = vardim
self.iteration = iteration
self.num_of_generations = num_of_generations
self.on_generation_finished = []
self.num_of_individuals = num_of_individuals
def __init__(self,
problem,
num_of_generations=1000,
num_of_individuals=100,
num_of_tour_particips=2,
tournament_prob=0.9,
crossover_param=2,
mutation_param=5):
self.utils = NSGA2Utils(problem, num_of_individuals,
num_of_tour_particips, tournament_prob,
crossover_param, mutation_param)
self.population = None
self.num_of_generations = num_of_generations
self.on_generation_finished = []
self.num_of_individuals = num_of_individuals
def init_population(self, portfolio_size, N, Q, Q_rfa, C):
"""
Parameters
----------
portfolio_size : int
number of risky assets
N : int
population size
Q : float
quantity limit for risky assets
C : int
cardinality limit
Q_rfa : float
quantity limit for RFA
Returns
-------
list
list of individuals
"""
solutions = []
for i in range(N):
if (np.all(self.initial_portfolio == None)):
solution = [
0 + (Q - 0) * np.random.random()
for j in range(portfolio_size)
]
solution.append(0 + (Q_rfa - 0) * np.random.random())
else:
solution = self.initial_portfolio
solutions.append(
NSGA2Utils.check_fix_solution(self, np.array(solution),
portfolio_size, C, Q, Q_rfa))
return solutions
x = i
children_test.append(x)
#try to run a single generation
newish_population = NSGAmethods.runNSGA_generation(pop_test, children_test,
testProblemParameters)
#stuff to set up the NSGA code's child creation proceedure
problem = Problem(num_of_variables=1,
objectives=[basic1, basic2],
variables_range=[(LB[0], UB[0])])
num_of_individuals = populationSize
num_of_tour_particips = 2
tournament_prob = 0.9
crossover_param = 2
mutation_param = 5
used_utils = NSGA2Utils(problem, num_of_individuals, num_of_tour_particips,
tournament_prob, crossover_param, mutation_param)
new_children_listOfIndividuals = used_utils.create_children(newish_population)
numChildren = len(new_children_listOfIndividuals)
#convert children a make them into a list of unevaluated arrays
#doing this is painful
#set up a loop and repeat until max generations
for GenCounter in range(0, MaxGenerations):
#step 1: combine parents and children into one population (which is technically done in runNSGA_generation)
#STEP 1A: Convert the children from a list of individuals (which store evaluation information)
#to a list of arrays (the way Gnowee store them) just containing their features
GnoweeChildList = []
for childNum in range(0, numChildren):
VarValue = new_children_listOfIndividuals[childNum].features[0]
GnoweeChildList.append(VarValue)