def setUp(self):
super().setUp()
self.seed = 7
aqua_globals.random_seed = self.seed
# Number training data samples
n_v = 5000
# Load data samples from log-normal distribution with mean=1 and standard deviation=1
m_u = 1
sigma = 1
self._real_data = aqua_globals.random.lognormal(mean=m_u, sigma=sigma, size=n_v)
# Set upper and lower data values as list of k
# min/max data values [[min_0,max_0],...,[min_k-1,max_k-1]]
self._bounds = [0., 3.]
# Set number of qubits per data dimension as list of k qubit values[#q_0,...,#q_k-1]
num_qubits = [2]
# Batch size
batch_size = 100
# Set number of training epochs
# num_epochs = 10
num_epochs = 5
# Initialize qGAN
self.qgan = QGAN(self._real_data,
self._bounds,
num_qubits,
batch_size,
num_epochs,
snapshot_dir=None)
self.qgan.seed = 7
# Set quantum instance to run the quantum generator
self.qi_statevector = QuantumInstance(backend=BasicAer.get_backend('statevector_simulator'),
seed_simulator=2,
seed_transpiler=2)
self.qi_qasm = QuantumInstance(backend=BasicAer.get_backend('qasm_simulator'),
shots=1000,
seed_simulator=2,
seed_transpiler=2)
# Set entangler map
entangler_map = [[0, 1]]
# Set an initial state for the generator circuit
init_dist = UniformDistribution(sum(num_qubits), low=self._bounds[0], high=self._bounds[1])
q = QuantumRegister(sum(num_qubits), name='q')
qc = QuantumCircuit(q)
init_dist.build(qc, q)
init_distribution = Custom(num_qubits=sum(num_qubits), circuit=qc)
# Set generator's initial parameters
init_params = aqua_globals.random.random(2 * sum(num_qubits)) * 2 * 1e-2
# Set variational form
var_form = RealAmplitudes(sum(num_qubits), reps=1, initial_state=init_distribution,
entanglement=entangler_map)
self.generator_circuit = UnivariateVariationalDistribution(sum(num_qubits), var_form,
init_params,
low=self._bounds[0],
high=self._bounds[1])
def test_ecev(self):
""" European Call Expected Value test """
bounds = np.array([0., 7.])
num_qubits = [3]
entangler_map = []
for i in range(sum(num_qubits)):
entangler_map.append([i, int(np.mod(i + 1, sum(num_qubits)))])
g_params = [
0.29399714, 0.38853322, 0.9557694, 0.07245791, 6.02626428,
0.13537225
]
# Set an initial state for the generator circuit
init_dist = NormalDistribution(int(sum(num_qubits)),
mu=1.,
sigma=1.,
low=bounds[0],
high=bounds[1])
init_distribution = np.sqrt(init_dist.probabilities)
init_distribution = Custom(num_qubits=sum(num_qubits),
state_vector=init_distribution)
var_form = TwoLocal(int(np.sum(num_qubits)),
'ry',
'cz',
reps=1,
initial_state=init_distribution,
entanglement=entangler_map)
uncertainty_model = UnivariateVariationalDistribution(int(
sum(num_qubits)),
var_form,
g_params,
low=bounds[0],
high=bounds[1])
strike_price = 2
c_approx = 0.25
european_call = EuropeanCallExpectedValue(uncertainty_model,
strike_price=strike_price,
c_approx=c_approx)
uncertainty_model.set_probabilities(
QuantumInstance(BasicAer.get_backend('statevector_simulator')))
algo = AmplitudeEstimation(5, european_call)
result = algo.run(
quantum_instance=BasicAer.get_backend('statevector_simulator'))
self.assertAlmostEqual(result['estimation'], 1.2580, places=4)
self.assertAlmostEqual(result['max_probability'], 0.8785, places=4)
def __init__(self,
bounds: np.ndarray,
num_qubits: List[int],
generator_circuit: Optional[
Union[UnivariateVariationalDistribution,
MultivariateVariationalDistribution,
QuantumCircuit]] = None,
init_params: Optional[Union[List[float], np.ndarray]] = None,
snapshot_dir: Optional[str] = None) -> None:
"""
Args:
bounds: k min/max data values [[min_1,max_1],...,[min_k,max_k]],
given input data dim k
num_qubits: k numbers of qubits to determine representation resolution,
i.e. n qubits enable the representation of 2**n values [n_1,..., n_k]
generator_circuit: a UnivariateVariationalDistribution for univariate data,
a MultivariateVariationalDistribution for multivariate data,
or a QuantumCircuit implementing the generator.
init_params: 1D numpy array or list, Initialization for
the generator's parameters.
snapshot_dir: str or None, if not None save the optimizer's parameter after every
update step to the given directory
Raises:
AquaError: Set multivariate variational distribution to represent multivariate data
"""
super().__init__()
self._bounds = bounds
self._num_qubits = num_qubits
self.generator_circuit = generator_circuit
if self.generator_circuit is None:
entangler_map = []
if np.sum(num_qubits) > 2:
for i in range(int(np.sum(num_qubits))):
entangler_map.append(
[i, int(np.mod(i + 1, np.sum(num_qubits)))])
else:
if np.sum(num_qubits) > 1:
entangler_map.append([0, 1])
if len(num_qubits) > 1:
num_qubits = list(map(int, num_qubits))
low = bounds[:, 0].tolist()
high = bounds[:, 1].tolist()
init_dist = MultivariateUniformDistribution(num_qubits,
low=low,
high=high)
q = QuantumRegister(sum(num_qubits))
qc = QuantumCircuit(q)
init_dist.build(qc, q)
init_distribution = Custom(num_qubits=sum(num_qubits),
circuit=qc)
# Set variational form
var_form = RY(sum(num_qubits),
depth=1,
initial_state=init_distribution,
entangler_map=entangler_map,
entanglement_gate='cz')
if init_params is None:
init_params = aqua_globals.random.rand(
var_form.num_parameters) * 2 * 1e-2
# Set generator circuit
self.generator_circuit = MultivariateVariationalDistribution(
num_qubits, var_form, init_params, low=low, high=high)
else:
init_dist = UniformDistribution(sum(num_qubits),
low=bounds[0],
high=bounds[1])
q = QuantumRegister(sum(num_qubits), name='q')
qc = QuantumCircuit(q)
init_dist.build(qc, q)
init_distribution = Custom(num_qubits=sum(num_qubits),
circuit=qc)
var_form = RY(sum(num_qubits),
depth=1,
initial_state=init_distribution,
entangler_map=entangler_map,
entanglement_gate='cz')
if init_params is None:
init_params = aqua_globals.random.rand(
var_form.num_parameters) * 2 * 1e-2
# Set generator circuit
self.generator_circuit = UnivariateVariationalDistribution(
int(np.sum(num_qubits)),
var_form,
init_params,
low=bounds[0],
high=bounds[1])
if len(num_qubits) > 1:
if isinstance(self.generator_circuit,
MultivariateVariationalDistribution):
pass
else:
raise AquaError('Set multivariate variational distribution '
'to represent multivariate data')
else:
if isinstance(self.generator_circuit,
UnivariateVariationalDistribution):
pass
else:
raise AquaError('Set univariate variational distribution '
'to represent univariate data')
# Set optimizer for updating the generator network
self._optimizer = ADAM(maxiter=1,
tol=1e-6,
lr=1e-3,
beta_1=0.7,
beta_2=0.99,
noise_factor=1e-6,
eps=1e-6,
amsgrad=True,
snapshot_dir=snapshot_dir)
if np.ndim(self._bounds) == 1:
bounds = np.reshape(self._bounds, (1, len(self._bounds)))
else:
bounds = self._bounds
for j, prec in enumerate(self._num_qubits):
# prepare data grid for dim j
grid = np.linspace(bounds[j, 0], bounds[j, 1], (2**prec))
if j == 0:
if len(self._num_qubits) > 1:
self._data_grid = [grid]
else:
self._data_grid = grid
self._grid_elements = grid
elif j == 1:
self._data_grid.append(grid)
temp = []
for g_e in self._grid_elements:
for g in grid:
temp0 = [g_e]
temp0.append(g)
temp.append(temp0)
self._grid_elements = temp
else:
self._data_grid.append(grid)
temp = []
for g_e in self._grid_elements:
for g in grid:
temp0 = deepcopy(g_e)
temp0.append(g)
temp.append(temp0)
self._grid_elements = deepcopy(temp)
self._data_grid = np.array(self._data_grid)
self._shots = None
self._discriminator = None
self._ret = {}
def setUp(self):
super().setUp()
# Number training data samples
N = 5000
# Load data samples from log-normal distribution with mean=1 and standard deviation=1
mu = 1
sigma = 1
self._real_data = np.random.lognormal(mean=mu, sigma=sigma, size=N)
# Set the data resolution
# Set upper and lower data values as list of k min/max data values [[min_0,max_0],...,[min_k-1,max_k-1]]
self._bounds = [0., 3.]
# Set number of qubits per data dimension as list of k qubit values[#q_0,...,#q_k-1]
num_qubits = [2]
# Batch size
batch_size = 100
# Set number of training epochs
num_epochs = 10
self._params_torch = {
'algorithm': {
'name': 'QGAN',
'num_qubits': num_qubits,
'batch_size': batch_size,
'num_epochs': num_epochs
},
'problem': {
'name': 'distribution_learning_loading',
'random_seed': 7
},
'generative_network': {
'name': 'QuantumGenerator',
'bounds': self._bounds,
'num_qubits': num_qubits,
'init_params': None,
'snapshot_dir': None
},
'discriminative_network': {
'name': 'PytorchDiscriminator',
'n_features': len(num_qubits)
}
}
self._params_numpy = {
'algorithm': {
'name': 'QGAN',
'num_qubits': num_qubits,
'batch_size': batch_size,
'num_epochs': num_epochs
},
'problem': {
'name': 'distribution_learning_loading',
'random_seed': 7
},
'generative_network': {
'name': 'QuantumGenerator',
'bounds': self._bounds,
'num_qubits': num_qubits,
'init_params': None,
'snapshot_dir': None
},
'discriminative_network': {
'name': 'NumpyDiscriminator',
'n_features': len(num_qubits)
}
}
# Initialize qGAN
self.qgan = QGAN(self._real_data,
self._bounds,
num_qubits,
batch_size,
num_epochs,
snapshot_dir=None)
self.qgan.seed = 7
# Set quantum instance to run the quantum generator
self.quantum_instance_statevector = QuantumInstance(
backend=BasicAer.get_backend('statevector_simulator'),
circuit_caching=False,
seed_simulator=2,
seed_transpiler=2)
self.quantum_instance_qasm = QuantumInstance(
backend=BasicAer.get_backend('qasm_simulator'),
shots=1000,
circuit_caching=False,
seed_simulator=2,
seed_transpiler=2)
# Set entangler map
entangler_map = [[0, 1]]
# Set an initial state for the generator circuit
init_dist = UniformDistribution(sum(num_qubits),
low=self._bounds[0],
high=self._bounds[1])
q = QuantumRegister(sum(num_qubits), name='q')
qc = QuantumCircuit(q)
init_dist.build(qc, q)
init_distribution = Custom(num_qubits=sum(num_qubits), circuit=qc)
# Set variational form
var_form = RY(sum(num_qubits),
depth=1,
initial_state=init_distribution,
entangler_map=entangler_map,
entanglement_gate='cz')
# Set generator's initial parameters
init_params = aqua_globals.random.rand(
var_form._num_parameters) * 2 * 1e-2
# Set generator circuit
g_circuit = UnivariateVariationalDistribution(sum(num_qubits),
var_form,
init_params,
low=self._bounds[0],
high=self._bounds[1])
# initial_distribution=init_distribution,
# Set quantum generator
self.qgan.set_generator(generator_circuit=g_circuit)
quantum_instance = QuantumInstance(
backend=BasicAer.get_backend('statevector_simulator'))
print("quantum_instance set")
# Set entangler map
entangler_map = [[0, 1]]
# Set an initial state for the generator circuit
init_dist = UniformDistribution(sum(num_qubits), low=bounds[0], high=bounds[1])
q = QuantumRegister(sum(num_qubits), name='q')
qc = QuantumCircuit(q)
init_dist.build(qc, q)
init_distribution = Custom(num_qubits=sum(num_qubits), circuit=qc)
var_form = TwoLocal(int(np.sum(num_qubits)), 'ry', 'cz', entanglement=entangler_map,
reps=1, initial_state=init_distribution)
# Set generator's initial parameters
init_params = aqua_globals.random.rand(
var_form.num_parameters_settable) * 2 * np.pi
# Set generator circuit
g_circuit = UnivariateVariationalDistribution(int(sum(num_qubits)), var_form, init_params,
low=bounds[0], high=bounds[1])
print("g_circuit set")
# Set quantum generator
qgan.set_generator(generator_circuit=g_circuit)
# Set classical discriminator neural network
discriminator = NumPyDiscriminator(len(num_qubits))
qgan.set_discriminator(discriminator)
def test_ecev(self, use_circuits):
""" European Call Expected Value test """
if not use_circuits:
# ignore deprecation warnings from the deprecation of VariationalForm as input for
# the univariate variational distribution
warnings.filterwarnings("ignore", category=DeprecationWarning)
bounds = np.array([0., 7.])
num_qubits = [3]
entangler_map = []
for i in range(sum(num_qubits)):
entangler_map.append([i, int(np.mod(i + 1, sum(num_qubits)))])
g_params = [
0.29399714, 0.38853322, 0.9557694, 0.07245791, 6.02626428,
0.13537225
]
# Set an initial state for the generator circuit
init_dist = NormalDistribution(int(sum(num_qubits)),
mu=1.,
sigma=1.,
low=bounds[0],
high=bounds[1])
init_distribution = np.sqrt(init_dist.probabilities)
init_distribution = Custom(num_qubits=sum(num_qubits),
state_vector=init_distribution)
var_form = RY(int(np.sum(num_qubits)),
depth=1,
initial_state=init_distribution,
entangler_map=entangler_map,
entanglement_gate='cz')
if use_circuits:
theta = ParameterVector('θ', var_form.num_parameters)
var_form = var_form.construct_circuit(theta)
uncertainty_model = UnivariateVariationalDistribution(int(
sum(num_qubits)),
var_form,
g_params,
low=bounds[0],
high=bounds[1])
if use_circuits:
uncertainty_model._var_form_params = theta
strike_price = 2
c_approx = 0.25
european_call = EuropeanCallExpectedValue(uncertainty_model,
strike_price=strike_price,
c_approx=c_approx)
uncertainty_model.set_probabilities(
QuantumInstance(BasicAer.get_backend('statevector_simulator')))
algo = AmplitudeEstimation(5, european_call)
result = algo.run(
quantum_instance=BasicAer.get_backend('statevector_simulator'))
self.assertAlmostEqual(result['estimation'], 1.2580, places=4)
self.assertAlmostEqual(result['max_probability'], 0.8785, places=4)
if not use_circuits:
warnings.filterwarnings(action="always",
category=DeprecationWarning)
def QGAN_method(kk, num_qubit, epochs, batch, bound, snap, data):
start = time.time()
real_data = data
# In[41]:
# Number training data samples
N = 1000
# Load data samples from log-normal distribution with mean=1 and standard deviation=1
mu = 1
sigma = 1
# real_data = np.random.lognormal(mean = mu, sigma=sigma, size=N)
# print(real_data)
# Set the data resolution
# Set upper and lower data values as list of k min/max data values [[min_0,max_0],...,[min_k-1,max_k-1]]
bounds = np.array([0, bound])
# Set number of qubits per data dimension as list of k qubit values[#q_0,...,#q_k-1]
num_qubits = [num_qubit]
k = kk
# In[52]:
# Set number of training epochs
# Note: The algorithm's runtime can be shortened by reducing the number of training epochs.
num_epochs = epochs
# Batch size
batch_size = batch
# Initialize qGAN
qgan = QGAN(real_data,
bounds,
num_qubits,
batch_size,
num_epochs,
snapshot_dir=snap)
qgan.seed = 1
# Set quantum instance to run the quantum generator
quantum_instance = QuantumInstance(
backend=BasicAer.get_backend('statevector_simulator'))
# Set entangler map
entangler_map = [[0, 1]]
# Set an initial state for the generator circuit
init_dist = UniformDistribution(sum(num_qubits),
low=bounds[0],
high=bounds[1])
q = QuantumRegister(sum(num_qubits), name='q')
qc = QuantumCircuit(q)
init_dist.build(qc, q)
init_distribution = Custom(num_qubits=sum(num_qubits), circuit=qc)
var_form = RY(int(np.sum(num_qubits)),
depth=k,
initial_state=init_distribution,
entangler_map=entangler_map,
entanglement_gate='cz')
# Set generator's initial parameters
init_params = aqua_globals.random.rand(
var_form._num_parameters) * 2 * np.pi
# Set generator circuit
g_circuit = UnivariateVariationalDistribution(int(sum(num_qubits)),
var_form,
init_params,
low=bounds[0],
high=bounds[1])
# Set quantum generator
qgan.set_generator(generator_circuit=g_circuit)
# Set classical discriminator neural network
discriminator = NumPyDiscriminator(len(num_qubits))
qgan.set_discriminator(discriminator)
# In[53]:
# Run qGAN
qgan.run(quantum_instance)
# Runtime
end = time.time()
print('qGAN training runtime: ', (end - start) / 60., ' min')
return qgan