class PITestCase(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([1])
self.X = init_random_uniform(self.X_lower, self.X_upper, 10)
self.Y = np.sin(self.X)
self.kernel = GPy.kern.RBF(input_dim=1)
self.model = GPyModel(self.kernel)
self.model.train(self.X, self.Y)
self.pi = PI(self.model,
X_upper=self.X_upper,
X_lower=self.X_lower)
def test_general_interface(self):
X_test = init_random_uniform(self.X_lower, self.X_upper, 10)
# Just check if PI is always greater equal than 0
a, dadx = self.pi(X_test, True)
assert len(a.shape) == 2
assert a.shape[0] == X_test.shape[0]
assert a.shape[1] == 1
assert np.all(a) >= 0.0
assert len(dadx.shape) == 2
assert dadx.shape[0] == X_test.shape[0]
assert dadx.shape[1] == X_test.shape[1]
def test_check_grads(self):
x_ = np.array([[np.random.rand()]])
assert check_grad(self.pi, lambda x: -self.pi(x, True)[1], x_) < 1e-5
class LogEITestCase(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([1])
self.X = np.random.rand(10)[:, np.newaxis]
self.Y = np.sin(self.X)
self.kernel = GPy.kern.RBF(input_dim=1)
self.model = GPyModel(self.kernel)
self.model.train(self.X, self.Y)
self.log_ei = LogEI(self.model,
X_upper=self.X_upper,
X_lower=self.X_lower)
def test_general_interface(self):
X_test = init_random_uniform(self.X_lower, self.X_upper, 10)
# a, dadx = self.log_ei(X_test, True)
a = self.log_ei(X_test)
assert len(a.shape) == 2
assert a.shape[0] == X_test.shape[0]
assert a.shape[1] == 1
class TestMaximizers2D(unittest.TestCase):
def setUp(self):
self.branin = Branin()
n_points = 5
self.X = np.random.rand(n_points, self.branin.n_dims)
self.X[:, 0] = self.X[:, 0].dot(self.branin.X_upper[0] - self.branin.X_lower[0]) + self.branin.X_lower[0]
self.X[:, 1] = self.X[:, 1].dot(self.branin.X_upper[1] - self.branin.X_lower[1]) + self.branin.X_lower[1]
self.Y = self.branin.evaluate(self.X)
kernel = GPy.kern.Matern52(input_dim=self.branin.n_dims)
self.model = GPyModel(kernel, optimize=True, noise_variance=1e-4, num_restarts=10)
self.model.train(self.X, self.Y)
self.acquisition_func = EI(
self.model,
X_upper=self.branin.X_upper,
X_lower=self.branin.X_lower,
compute_incumbent=compute_incumbent,
par=0.1,
)
def test_direct(self):
maximizer = Direct(self.acquisition_func, self.branin.X_lower, self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test_stochastic_local_search(self):
maximizer = StochasticLocalSearch(self.acquisition_func, self.branin.X_lower, self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test_cmaes(self):
maximizer = CMAES(self.acquisition_func, self.branin.X_lower, self.branin.X_upper)
x = maximizer.maximize(verbose=False)
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
class EITestCase(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([1])
self.X = init_random_uniform(self.X_lower, self.X_upper, 10)
self.Y = np.sin(self.X)
self.kernel = GPy.kern.RBF(input_dim=1)
self.model = GPyModel(self.kernel)
self.model.train(self.X, self.Y)
self.ei = EI(self.model, X_upper=self.X_upper, X_lower=self.X_lower)
def test_general_interface(self):
X_test = init_random_uniform(self.X_lower, self.X_upper, 10)
# Just check if EI is always greater equal than 0
a, dadx = self.ei(X_test, True)
assert len(a.shape) == 2
assert a.shape[0] == X_test.shape[0]
assert a.shape[1] == 1
assert np.all(a) >= 0.0
assert len(dadx.shape) == 2
assert dadx.shape[0] == X_test.shape[0]
assert dadx.shape[1] == X_test.shape[1]
def test_check_grads(self):
x_ = np.array([[np.random.rand()]])
assert check_grad(self.ei, lambda x: -self.ei(x, True)[1], x_) < 1e-5
class LCBTestCase(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([1])
self.X = init_random_uniform(self.X_lower, self.X_upper, 4)
self.Y = np.sin(self.X)
self.kernel = GPy.kern.RBF(input_dim=1)
self.model = GPyModel(self.kernel)
self.model.train(self.X, self.Y)
self.lcb = LCB(self.model,
X_upper=self.X_upper,
X_lower=self.X_lower)
def test_general_interface(self):
X_test = init_random_uniform(self.X_lower, self.X_upper, 10)
a, dadx = self.lcb(X_test, True)
assert len(a.shape) == 2
assert a.shape[0] == X_test.shape[0]
assert a.shape[1] == 1
assert len(dadx.shape) == 2
assert dadx.shape[0] == X_test.shape[0]
assert dadx.shape[1] == X_test.shape[1]
def test_check_grads(self):
x_ = np.array([[np.random.rand()]])
assert check_grad(self.lcb, lambda x: -self.lcb(x, True)[1], x_) < 1e-5
def train(self, X, Y, *args):
self.X = X
self.Y = Y
if X.size == 0 or Y.size == 0:
return
m = GPy.models.GPRegression(self.X, self.Y, self.kernel)
# Add exponential prior for the noise
m.likelihood.unconstrain()
m.likelihood.variance.set_prior(GPy.priors.Exponential(1))
m.likelihood.variance.constrain_positive()
if self.hmc is None:
self.hmc = GPy.inference.mcmc.hmc.HMC(m, stepsize=5e-2)
# Burnin
self.hmc.sample(num_samples=self.burnin)
else:
self.hmc.model = m
# Start the mcmc chain
self.mcmc_chain = self.hmc.sample(num_samples=(self.chain_length))
self.samples = self.mcmc_chain[list(range(0, self.chain_length, self.chain_length // self.n_hypers))]
self.models = []
for sample in self.samples:
# Instantiate a model for each hyperparam configuration
kernel = deepcopy(self.kernel)
for i in range(len(sample) - 1):
kernel.param_array[i] = sample[i]
model = GPyModel(kernel, noise_variance=sample[-1])
model.train(self.X, self.Y, optimize=False)
self.models.append(model)
class Test(unittest.TestCase):
def func(self, x):
return x ** x + 0.5
def setUp(self):
X = np.array([[0.0, 0.25, 0.75, 1.0]])
X = X.T
y = self.func(X)
k = GPy.kern.RBF(input_dim=1)
self.m = GPyModel(k, noise_variance=1e-3)
self.m.train(X, y)
self.X_lower = np.array([0])
self.X_upper = np.array([1.0])
def tearDown(self):
pass
def test_optimize_posterior_mean(self):
inc, inc_val = optimize_posterior_mean(self.m, self.X_lower, self.X_upper, with_gradients=True)
assert len(inc.shape) == 1
assert np.all(inc >= self.X_lower)
assert np.all(inc <= self.X_upper)
assert np.all(inc_val <= compute_incumbent(self.m)[1])
def test_optimize_posterior_mean_and_std(self):
inc, inc_val = optimize_posterior_mean_and_std(self.m, self.X_lower, self.X_upper, with_gradients=True)
assert len(inc.shape) == 1
assert np.all(inc >= self.X_lower)
assert np.all(inc <= self.X_upper)
assert np.all(inc_val <= compute_incumbent(self.m)[1])
class LCBTestCase(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([1])
self.X = init_random_uniform(self.X_lower, self.X_upper, 4)
self.Y = np.sin(self.X)
self.kernel = GPy.kern.RBF(input_dim=1)
self.model = GPyModel(self.kernel)
self.model.train(self.X, self.Y)
self.lcb = LCB(self.model, X_upper=self.X_upper, X_lower=self.X_lower)
def test_general_interface(self):
X_test = init_random_uniform(self.X_lower, self.X_upper, 10)
a, dadx = self.lcb(X_test, True)
assert len(a.shape) == 2
assert a.shape[0] == X_test.shape[0]
assert a.shape[1] == 1
assert len(dadx.shape) == 2
assert dadx.shape[0] == X_test.shape[0]
assert dadx.shape[1] == X_test.shape[1]
def test_check_grads(self):
x_ = np.array([[np.random.rand()]])
assert check_grad(self.lcb, lambda x: -self.lcb(x, True)[1], x_) < 1e-5
class TestMaximizers1D(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([6])
self.dims = 1
self.X = np.array([[1], [3.8], [0.9], [5.2], [3.4]])
self.X[:, 0] = self.X[:, 0].dot(self.X_upper[0] -
self.X_lower[0]) + self.X_lower[0]
self.Y = objective_function(self.X)
kernel = GPy.kern.Matern52(input_dim=self.dims)
self.model = GPyModel(kernel,
optimize=True,
noise_variance=1e-4,
num_restarts=10)
self.model.train(self.X, self.Y)
self.acquisition_func = EI(self.model,
X_upper=self.X_upper,
X_lower=self.X_lower,
par=0.1)
def test_direct(self):
maximizer = Direct(self.acquisition_func, self.X_lower, self.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.dims
assert np.all(x[:, 0] >= self.X_lower[0])
assert np.all(x[:, 0] <= self.X_upper[0])
assert np.all(x < self.X_upper)
def test_stochastic_local_search(self):
maximizer = StochasticLocalSearch(self.acquisition_func, self.X_lower,
self.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.dims
assert np.all(x[:, 0] >= self.X_lower[0])
assert np.all(x[:, 0] <= self.X_upper[0])
assert np.all(x < self.X_upper)
def test_grid_search(self):
maximizer = GridSearch(self.acquisition_func, self.X_lower,
self.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.dims
assert np.all(x[:, 0] >= self.X_lower[0])
assert np.all(x[:, 0] <= self.X_upper[0])
assert np.all(x < self.X_upper)
class TestMaximizers1D(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([6])
self.dims = 1
self.X = np.array([[1], [3.8], [0.9], [5.2], [3.4]])
self.X[:, 0] = self.X[:, 0].dot(self.X_upper[0] - self.X_lower[0]) + self.X_lower[0]
self.Y = objective_function(self.X)
kernel = GPy.kern.Matern52(input_dim=self.dims)
self.model = GPyModel(kernel, optimize=True,
noise_variance=1e-4, num_restarts=10)
self.model.train(self.X, self.Y)
self.acquisition_func = EI(self.model, X_upper=self.X_upper,
X_lower=self.X_lower,
par=0.1)
def test_direct(self):
maximizer = Direct(self.acquisition_func, self.X_lower, self.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.dims
assert np.all(x[:, 0] >= self.X_lower[0])
assert np.all(x[:, 0] <= self.X_upper[0])
assert np.all(x < self.X_upper)
def test_stochastic_local_search(self):
maximizer = StochasticLocalSearch(self.acquisition_func,
self.X_lower, self.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.dims
assert np.all(x[:, 0] >= self.X_lower[0])
assert np.all(x[:, 0] <= self.X_upper[0])
assert np.all(x < self.X_upper)
def test_grid_search(self):
maximizer = GridSearch(self.acquisition_func,
self.X_lower,
self.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.dims
assert np.all(x[:, 0] >= self.X_lower[0])
assert np.all(x[:, 0] <= self.X_upper[0])
assert np.all(x < self.X_upper)
class PITestCase1(unittest.TestCase):
def setUp(self):
self.x = np.array([[0.62971589], [0.63273273], [0.17867868], [0.17447447], [1.88558559]])
self.y = np.array([[-3.69925653], [-3.66221988], [-3.65560591], [-3.58907791], [-8.06925984]])
self.kernel = GPy.kern.RBF(input_dim=1, variance=30.1646253727, lengthscale=0.435343653946)
self.noise = 1e-20
self.model = GPyModel(self.kernel, noise_variance=self.noise, optimize=False)
self.model.train(self.x, self.y)
def test(self):
X_upper = np.array([2.1])
X_lower = np.array([-2.1])
x_test = np.array([[1.7], [2.0]])
pi_estimator = PI(self.model, X_lower, X_upper, compute_incumbent)
assert pi_estimator(x_test[0, np.newaxis])[0] > 0.0
assert pi_estimator(x_test[1, np.newaxis])[0] > 0.0
self.assertAlmostEqual(pi_estimator(self.x[-1, np.newaxis])[0], 0.0, delta=10E-5)
class LogEITestCase(unittest.TestCase):
def setUp(self):
self.X_lower = np.array([0])
self.X_upper = np.array([1])
self.X = np.random.rand(10)[:, np.newaxis]
self.Y = np.sin(self.X)
self.kernel = GPy.kern.RBF(input_dim=1)
self.model = GPyModel(self.kernel)
self.model.train(self.X, self.Y)
self.log_ei = LogEI(self.model,
X_upper=self.X_upper,
X_lower=self.X_lower)
def test_general_interface(self):
X_test = init_random_uniform(self.X_lower, self.X_upper, 10)
# a, dadx = self.log_ei(X_test, True)
a = self.log_ei(X_test)
assert len(a.shape) == 2
assert a.shape[0] == X_test.shape[0]
assert a.shape[1] == 1
class TestMaximizers2D(unittest.TestCase):
def setUp(self):
self.branin = Branin()
n_points = 5
rng = np.random.RandomState(42)
self.X = init_random_uniform(self.branin.X_lower,
self.branin.X_upper,
n_points,
rng=rng)
self.Y = self.branin.evaluate(self.X)
kernel = GPy.kern.Matern52(input_dim=self.branin.n_dims)
self.model = GPyModel(kernel, optimize=True,
noise_variance=1e-4,
num_restarts=10)
self.model.train(self.X, self.Y)
self.acquisition_func = EI(self.model,
X_upper=self.branin.X_upper,
X_lower=self.branin.X_lower,
par=0.1)
def test_direct(self):
maximizer = Direct(self.acquisition_func,
self.branin.X_lower,
self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test_stochastic_local_search(self):
maximizer = StochasticLocalSearch(self.acquisition_func,
self.branin.X_lower,
self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test_cmaes(self):
maximizer = CMAES(self.acquisition_func,
self.branin.X_lower,
self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
class TestMaximizers2D(unittest.TestCase):
def setUp(self):
self.branin = Branin()
n_points = 5
rng = np.random.RandomState(42)
self.X = init_random_uniform(self.branin.X_lower,
self.branin.X_upper,
n_points,
rng=rng)
self.Y = self.branin.evaluate(self.X)
kernel = GPy.kern.Matern52(input_dim=self.branin.n_dims)
self.model = GPyModel(kernel,
optimize=True,
noise_variance=1e-4,
num_restarts=10)
self.model.train(self.X, self.Y)
self.acquisition_func = EI(self.model,
X_upper=self.branin.X_upper,
X_lower=self.branin.X_lower,
par=0.1)
def test_direct(self):
maximizer = Direct(self.acquisition_func, self.branin.X_lower,
self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test_stochastic_local_search(self):
maximizer = StochasticLocalSearch(self.acquisition_func,
self.branin.X_lower,
self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test_cmaes(self):
maximizer = CMAES(self.acquisition_func, self.branin.X_lower,
self.branin.X_upper)
x = maximizer.maximize()
assert x.shape[0] == 1
assert x.shape[1] == self.branin.n_dims
assert np.all(x[:, 0] >= self.branin.X_lower[0])
assert np.all(x[:, 1] >= self.branin.X_lower[1])
assert np.all(x[:, 0] <= self.branin.X_upper[0])
assert np.all(x[:, 1] <= self.branin.X_upper[1])
assert np.all(x < self.branin.X_upper)
def test(self):
X_lower = np.array([0])
X_upper = np.array([1])
X = init_random_uniform(X_lower, X_upper, 10)
Y = np.sin(X)
kernel = GPy.kern.Matern52(input_dim=1)
model = GPyModel(kernel)
model.train(X, Y)
x_test = init_random_uniform(X_lower, X_upper, 3)
# Shape matching predict
m, v = model.predict(x_test, full_cov=True)
assert len(m.shape) == 2
assert m.shape[0] == x_test.shape[0]
assert m.shape[1] == 1
assert len(v.shape) == 2
assert v.shape[0] == x_test.shape[0]
assert v.shape[1] == x_test.shape[0]
# Check gradients
dm, dv = model.predictive_gradients(x_test)
assert len(dm.shape) == 2
assert dm.shape[0] == x_test.shape[0]
assert dm.shape[1] == x_test.shape[1]
assert len(dv.shape) == 2
assert dv.shape[0] == x_test.shape[0]
assert dv.shape[1] == 1
# Shape matching function sampling
x_ = np.linspace(X_lower, X_upper, 10)
x_ = x_[:, np.newaxis]
funcs = model.sample_functions(x_, n_funcs=2)
assert len(funcs.shape) == 2
assert funcs.shape[0] == 2
assert funcs.shape[1] == x_.shape[0]
# Shape matching predict variance
x_test2 = np.array([np.random.rand(1)])
x_test1 = np.random.rand(10)[:, np.newaxis]
var = model.predict_variance(x_test1, x_test2)
assert len(var.shape) == 2
assert var.shape[0] == x_test1.shape[0]
assert var.shape[1] == 1
# Check compatibility with all acquisition functions
acq_func = EI(model,
X_upper=X_upper,
X_lower=X_lower)
acq_func.update(model)
acq_func(x_test)
acq_func = PI(model,
X_upper=X_upper,
X_lower=X_lower)
acq_func.update(model)
acq_func(x_test)
acq_func = LCB(model,
X_upper=X_upper,
X_lower=X_lower)
acq_func.update(model)
acq_func(x_test)
acq_func = InformationGain(model,
X_upper=X_upper,
X_lower=X_lower)
acq_func.update(model)
acq_func(x_test)
# Check compatibility with all incumbent estimation methods
rec = BestObservation(model, X_lower, X_upper)
inc, inc_val = rec.estimate_incumbent(None)
assert len(inc.shape) == 2
assert inc.shape[0] == 1
assert inc.shape[1] == X_upper.shape[0]
assert len(inc_val.shape) == 2
assert inc_val.shape[0] == 1
assert inc_val.shape[1] == 1
rec = PosteriorMeanOptimization(model, X_lower, X_upper)
startpoints = init_random_uniform(X_lower, X_upper, 4)
inc, inc_val = rec.estimate_incumbent(startpoints)
assert len(inc.shape) == 2
assert inc.shape[0] == 1
assert inc.shape[1] == X_upper.shape[0]
assert len(inc_val.shape) == 2
assert inc_val.shape[0] == 1
assert inc_val.shape[1] == 1
rec = PosteriorMeanAndStdOptimization(model, X_lower, X_upper)
startpoints = init_random_uniform(X_lower, X_upper, 4)
inc, inc_val = rec.estimate_incumbent(startpoints)
assert len(inc.shape) == 2
assert inc.shape[0] == 1
assert inc.shape[1] == X_upper.shape[0]
assert len(inc_val.shape) == 2
assert inc_val.shape[0] == 1
assert inc_val.shape[1] == 1