def test_BO(dim, obj_fun, ftarget, max_FEs, lb, ub, logfile):
sys.path.insert(0, '../')
from bayes_optim import AnnealingBO, BO, ContinuousSpace
from bayes_optim.Surrogate import GaussianProcess, trend
space = ContinuousSpace(list(zip(lb, ub)))
mean = trend.constant_trend(dim,
beta=None) # equivalent to Ordinary Kriging
thetaL = 1e-10 * (ub - lb) * np.ones(dim)
thetaU = 10 * (ub - lb) * np.ones(dim)
theta0 = np.random.rand(dim) * (thetaU - thetaL) + thetaL
model = GaussianProcess(mean=mean,
corr='matern',
theta0=theta0,
thetaL=thetaL,
thetaU=thetaU,
noise_estim=False,
nugget=1e-6,
optimizer='BFGS',
wait_iter=5,
random_start=5 * dim,
likelihood='concentrated',
eval_budget=100 * dim)
return BO(search_space=space,
obj_fun=obj_fun,
model=model,
DoE_size=dim * 5,
max_FEs=max_FEs,
verbose=False,
n_point=1,
minimize=True,
ftarget=ftarget,
logger=logfile)
def test_mix_space():
dim_r = 2 # dimension of the real values
def obj_fun(x):
x_r = np.array([x['continuous_%d' % i] for i in range(dim_r)])
x_i = x['ordinal']
x_d = x['nominal']
_ = 0 if x_d == 'OK' else 1
return np.sum(x_r**2) + abs(x_i - 10) / 123. + _ * 2
search_space = ContinuousSpace([-5, 5], var_name='continuous') * dim_r + \
OrdinalSpace([5, 15], var_name='ordinal') + \
NominalSpace(['OK', 'A', 'B', 'C', 'D', 'E', 'F', 'G'], var_name='nominal')
model = RandomForest(levels=search_space.levels)
opt = ParallelBO(
search_space=search_space,
obj_fun=obj_fun,
model=model,
max_FEs=6,
DoE_size=3, # the initial DoE size
eval_type='dict',
acquisition_fun='MGFI',
acquisition_par={'t': 2},
n_job=3, # number of processes
n_point=3, # number of the candidate solution proposed in each iteration
verbose=True # turn this off, if you prefer no output
)
xopt, fopt, stop_dict = opt.run()
print('xopt: {}'.format(xopt))
print('fopt: {}'.format(fopt))
print('stop criteria: {}'.format(stop_dict))
def test_sampling():
C = ContinuousSpace([-5, 5]) * 3
I = OrdinalSpace([[-100, 100], [-5, 5]], 'heihei')
N = NominalSpace([['OK', 'A', 'B', 'C', 'D', 'E', 'A']] * 2, ['x', 'y'])
S = N + I + C
S.sampling(5, method='LHS')
S.sampling(5, method='uniform')
def test_scale():
C = ContinuousSpace([1, 5], scale='log')
assert C.bounds[0][0] == 0
C = ContinuousSpace([0.5, 0.8], scale='logit')
assert C.bounds[0][0] == 0
C = ContinuousSpace([-1, 1], scale='bilog')
assert C.bounds[0][0] == -np.log(2)
assert C.bounds[0][1] == np.log(2)
C = ContinuousSpace([-1, 1], scale='bilog') * 2
X = np.array([-np.log(2), np.log(2)])
a = C.to_linear_scale(X)
assert all(a == np.array([-1, 1]))
def test_warm_data_with_RF():
space = ContinuousSpace([-10, 10]) * 2 + \
OrdinalSpace([5, 15]) + \
NominalSpace(['OK', 'A', 'B', 'C', 'D', 'E', 'F', 'G'])
X = space.sampling(10)
y = [obj_fun(x) for x in X]
model = RandomForest(levels=space.levels)
opt = BO(search_space=space,
obj_fun=obj_fun,
model=model,
minimize=True,
eval_type='list',
max_FEs=10,
verbose=True,
acquisition_fun='EI',
warm_data=(X, y))
opt.run()
assert opt.data.shape[0] == 20
def test_ProductSpace():
C = ContinuousSpace([-5, 5], precision=1) * 3 # product of the same space
I = OrdinalSpace([[-100, 100], [-5, 5]], 'heihei')
N = NominalSpace([['OK', 'A', 'B', 'C', 'D', 'E', 'A']] * 2, ['x', 'y'])
space = C + C + C
print(space.sampling(2))
# cartesian product of heterogeneous spaces
space = C + I + N
print(space.sampling(10))
print(space.bounds)
print(space.var_name)
print(space.var_type)
print((C * 2).var_name)
print((N * 3).sampling(2))
C = ContinuousSpace([[0, 1]] * 2, var_name='weight')
print(C.var_name)
def test_precision():
C = ContinuousSpace([-5, 5], precision=2) * 3
X = C.round(C.sampling(1, method='LHS'))
X = [re.sub(r'^-?\d+\.(\d+)$', r'\1', str(_)) for _ in X[0]]
assert all([len(x) <= 2 for x in X])
X = C.round(C.sampling(1, method='uniform'))
X = [re.sub(r'^-?\d+\.(\d+)$', r'\1', str(_)) for _ in X[0]]
assert all([len(x) <= 2 for x in X])
X = np.random.rand(2, 3) * 10 - 5
assert isinstance(C.round(X), np.ndarray)
X_ = C.round(X.tolist())
assert isinstance(X_, list)
assert isinstance(X_[0], list)
assert np.all(np.array(X_) == C.round(X))
def test_BO_constraints():
search_space = OrdinalSpace([1, 10], var_name='mu') + \
OrdinalSpace([1, 10], var_name='lambda') + \
ContinuousSpace([0, 1], var_name='pc') + \
ContinuousSpace([0.005, 0.5], var_name='p')
model = RandomForest(levels=search_space.levels)
xopt, _, __ = BO(search_space=search_space,
obj_fun=obj_func,
ineq_fun=g,
model=model,
max_FEs=30,
DoE_size=3,
eval_type='dict',
acquisition_fun='MGFI',
acquisition_par={
't': 2
},
n_job=1,
n_point=1,
verbose=True).run()
assert isinstance(xopt, dict)
assert all(np.array(g(xopt)) <= 0)
def test_pickling2():
dim = 5
lb, ub = -1, 5
def fitness(x):
x = np.asarray(x)
return np.sum(x**2)
space = ContinuousSpace([lb, ub]) * dim
mean = trend.constant_trend(dim, beta=None)
thetaL = 1e-10 * (ub - lb) * np.ones(dim)
thetaU = 10 * (ub - lb) * np.ones(dim)
theta0 = np.random.rand(dim) * (thetaU - thetaL) + thetaL
model = GaussianProcess(mean=mean,
corr='squared_exponential',
theta0=theta0,
thetaL=thetaL,
thetaU=thetaU,
nugget=0,
noise_estim=False,
optimizer='BFGS',
wait_iter=3,
random_start=dim,
likelihood='concentrated',
eval_budget=100 * dim)
opt = BO(search_space=space,
obj_fun=fitness,
model=model,
DoE_size=5,
max_FEs=10,
verbose=True,
n_point=1,
logger='log')
opt.save('test')
opt = BO.load('test')
print(opt.run())
os.remove('test')
os.remove('log')
def test_warm_data_with_GPR():
dim = 2
lb, ub = -5, 5
def fitness(x):
x = np.asarray(x)
return np.sum(x**2)
X = np.random.rand(5, dim) * (ub - lb) + lb
y = [fitness(x) for x in X]
space = ContinuousSpace([lb, ub]) * dim
thetaL = 1e-10 * (ub - lb) * np.ones(dim)
thetaU = 10 * (ub - lb) * np.ones(dim)
theta0 = np.random.rand(dim) * (thetaU - thetaL) + thetaL
model = GaussianProcess(theta0=theta0,
thetaL=thetaL,
thetaU=thetaU,
nugget=0,
noise_estim=False,
optimizer='BFGS',
wait_iter=3,
random_start=dim,
likelihood='concentrated',
eval_budget=100 * dim)
opt = BO(search_space=space,
obj_fun=fitness,
model=model,
warm_data=(X, y),
max_FEs=10,
verbose=True,
n_point=1)
assert np.all(np.asarray(opt.data) == np.asarray(opt.warm_data))
assert opt.model.is_fitted
opt.run()
sys.path.insert(0, '../')
from bayes_optim import BO, ContinuousSpace
from bayes_optim.Surrogate import GaussianProcess, trend
np.random.seed(123)
dim = 5
lb, ub = -1, 5
def fitness(x):
x = np.asarray(x)
return np.sum(x**2)
space = ContinuousSpace([lb, ub]) * dim
mean = trend.constant_trend(dim, beta=None)
thetaL = 1e-10 * (ub - lb) * np.ones(dim)
thetaU = 10 * (ub - lb) * np.ones(dim)
theta0 = np.random.rand(dim) * (thetaU - thetaL) + thetaL
model = GaussianProcess(theta0=theta0,
thetaL=thetaL,
thetaU=thetaU,
nugget=0,
noise_estim=False,
optimizer='BFGS',
wait_iter=3,
random_start=dim,
likelihood='concentrated',
def test_MIES():
def obj_fun(X):
v = []
for x in X:
x_r = np.array([x['real_%d' % i] for i in range(2)])
x_i = np.array([x['int_%d' % i] for i in range(2)])
x_d = np.array([x['cat_%d' % i] for i in range(2)])
v.append(
np.sum(x_r ** 2) + \
np.sum(np.abs(x_i - 10) / 123.) + \
np.sum(x_d == 'OK')
)
return v
search_space = ContinuousSpace([-5, 5], var_name='real') * 2 + \
OrdinalSpace([5, 15], var_name='int') * 2 + \
NominalSpace(
['OK', 'A', 'B', 'C', 'D', 'E', 'F', 'G'],
var_name='cat'
) * 2
opt = MIES(
search_space=search_space,
obj_func=obj_fun,
max_eval=100,
eval_type='dict',
verbose=True # turn this off, if you prefer no output
)
xopt, fopt, stop_dict = opt.optimize()
print('xopt: {}'.format(xopt))
print('fopt: {}'.format(fopt))
print('stop criteria: {}'.format(stop_dict))
# def test_multi_acquisition():
# dim_r = 2 # dimension of the real values
# def obj_fun(x):
# x_r = np.array([x['continuous_%d'%i] for i in range(dim_r)])
# x_i = x['ordinal']
# x_d = x['nominal']
# _ = 0 if x_d == 'OK' else 1
# return np.sum(x_r ** 2) + abs(x_i - 10) / 123. + _ * 2
# search_space = ContinuousSpace([-5, 5], var_name='continuous') * dim_r + \
# OrdinalSpace([5, 15], var_name='ordinal') + \
# NominalSpace(['OK', 'A', 'B', 'C', 'D', 'E', 'F', 'G'], var_name='nominal')
# model = RandomForest(levels=search_space.levels)
# opt = MultiAcquisitionBO(
# search_space=search_space,
# obj_fun=obj_fun,
# model=model,
# max_FEs=8,
# DoE_size=4, # the initial DoE size
# eval_type='dict',
# n_job=4, # number of processes
# n_point=4, # number of the candidate solution proposed in each iteration
# verbose=True # turn this off, if you prefer no output
# )
# xopt, fopt, stop_dict = opt.run()
# print('xopt: {}'.format(xopt))
# print('fopt: {}'.format(fopt))
# print('stop criteria: {}'.format(stop_dict))
_ = 0 if x_d == 'OK' else 1
return np.sum(x_r**2) + abs(x_i - 10) / 123. + _ * 2
def obj_fun(x):
x_r = np.array([x['continuous_%d' % i] for i in range(dim_r)])
x_i = x['ordinal']
x_d = x['nominal']
_ = 0 if x_d == 'OK' else 1
return np.sum(x_r**2) + abs(x_i - 10) / 123. + _ * 2
# Continuous variables can be specified as follows:
# a 2-D variable in [-5, 5]^2
# for 2 variables, the naming scheme is continuous0, continuous1
C = ContinuousSpace([-5, 5], var_name='continuous') * dim_r
# Equivalently, you can also use
# C = ContinuousSpace([[-5, 5]]] * dim)
# The general usage is:
# ContinuousSpace([[lb_1, ub_1], [lb_2, ub_2], ..., [lb_n, ub_n]])
# Integer (ordinal) variables can be specified as follows:
# The domain of integer variables can be given as with continuous ones
# var_name is optional
I = OrdinalSpace([5, 15], var_name='ordinal')
# Discrete (nominal) variables can be specified as follows:
# No lb, ub... a list of categories instead
N = NominalSpace(['OK', 'A', 'B', 'C', 'D', 'E', 'F', 'G'], var_name='nominal')
from bayes_optim import ContinuousSpace, OrdinalSpace, NominalSpace
from bayes_optim.Surrogate import RandomForest
from bayes_optim.Extension import MultiAcquisitionBO
dim_r = 2 # dimension of the real values
def obj_fun(x):
x_r = np.array([x['continuous_%d' % i] for i in range(dim_r)])
x_i = x['ordinal']
x_d = x['nominal']
_ = 0 if x_d == 'OK' else 1
return np.sum(x_r**2) + abs(x_i - 10) / 123. + _ * 2
search_space = ContinuousSpace([-5, 5], var_name='continuous') * dim_r + \
OrdinalSpace([5, 15], var_name='ordinal') + \
NominalSpace(['OK', 'A', 'B', 'C', 'D', 'E', 'F', 'G'], var_name='nominal')
model = RandomForest(levels=search_space.levels)
opt = MultiAcquisitionBO(
search_space=search_space,
obj_fun=obj_fun,
model=model,
max_FEs=40,
DoE_size=4, # the initial DoE size
eval_type='dict',
n_job=4, # number of processes
n_point=4, # number of the candidate solution proposed in each iteration
verbose=True # turn this off, if you prefer no output
)
