def init_params(self, params, algo_input):
algo_params = params.get(QuantumAlgorithm.SECTION_KEY_ALGORITHM)
override_spsa_params = algo_params.get('override_SPSA_params')
# Set up optimizer
opt_params = params.get(QuantumAlgorithm.SECTION_KEY_OPTIMIZER)
optimizer = get_optimizer_instance(opt_params['name'])
# If SPSA then override SPSA params as reqd to our predetermined values
if opt_params['name'] == 'SPSA' and override_spsa_params:
opt_params['c0'] = 4.0
opt_params['c1'] = 0.1
opt_params['c2'] = 0.602
opt_params['c3'] = 0.101
opt_params['c4'] = 0.0
opt_params['skip_calibration'] = True
optimizer.init_params(opt_params)
# Set up variational form
fea_map_params = params.get(QuantumAlgorithm.SECTION_KEY_FEATURE_MAP)
num_qubits = get_feature_dimension(algo_input.training_dataset)
fea_map_params['num_qubits'] = num_qubits
feature_map = get_feature_map_instance(fea_map_params['name'])
feature_map.init_params(fea_map_params)
# Set up variational form
var_form_params = params.get(QuantumAlgorithm.SECTION_KEY_VAR_FORM)
var_form_params['num_qubits'] = num_qubits
var_form = get_variational_form_instance(var_form_params['name'])
var_form.init_params(var_form_params)
self.init_args(algo_input.training_dataset, algo_input.test_dataset,
algo_input.datapoints, optimizer, feature_map, var_form)
def init_params(self, params, algo_input):
# Set up optimizer
opt_params = params.get(QuantumAlgorithm.SECTION_KEY_OPTIMIZER)
optimizer = get_optimizer_instance(opt_params['name'])
# hard-coded params if SPSA is used.
if opt_params['name'] == 'SPSA' and opt_params['parameters'] is None:
opt_params['parameters'] = np.asarray(
[4.0, 0.1, 0.602, 0.101, 0.0])
optimizer.init_params(opt_params)
# Set up variational form
fea_map_params = params.get(QuantumAlgorithm.SECTION_KEY_FEATURE_MAP)
num_qubits = self._auto_detect_qubitnum(algo_input.training_dataset)
fea_map_params['num_qubits'] = num_qubits
feature_map = get_feature_map_instance(fea_map_params['name'])
feature_map.init_params(fea_map_params)
# Set up variational form
var_form_params = params.get(QuantumAlgorithm.SECTION_KEY_VAR_FORM)
var_form_params['num_qubits'] = num_qubits
var_form = get_variational_form_instance(var_form_params['name'])
var_form.init_params(var_form_params)
self.init_args(algo_input.training_dataset, algo_input.test_dataset,
algo_input.datapoints, optimizer, feature_map, var_form)
def init_params(self, params, algo_input):
if algo_input.training_dataset is None:
raise AlgorithmError("Training dataset is required.")
fea_map_params = params.get(QuantumAlgorithm.SECTION_KEY_FEATURE_MAP)
feature_map = get_feature_map_instance(fea_map_params['name'])
num_qubits = get_feature_dimension(algo_input.training_dataset)
fea_map_params['num_qubits'] = num_qubits
feature_map.init_params(fea_map_params)
is_multiclass = get_num_classes(algo_input.training_dataset) > 2
if is_multiclass:
multicls_ext_params = params.get(
QuantumAlgorithm.SECTION_KEY_MULTICLASS_EXTENSION)
multiclass_extension = get_multiclass_extension_instance(
multicls_ext_params['name'])
# we need to set this explicitly for quantum version
multicls_ext_params['params'] = [feature_map, self]
multicls_ext_params['estimator_cls'] = QKernalSVM_Estimator
multiclass_extension.init_params(multicls_ext_params)
logger.info("Multiclass classifcation algo:" +
multicls_ext_params['name'])
else:
logger.warning(
"Only two classes in the dataset, use binary classifer"
" and ignore all options of multiclass_extension")
multiclass_extension = None
self.init_args(algo_input.training_dataset, algo_input.test_dataset,
algo_input.datapoints, feature_map,
multiclass_extension)
def test_svm_qkernel_binary_directly(self):
svm = get_algorithm_instance("QSVM.Kernel")
svm.random_seed = self.random_seed
svm.setup_quantum_backend(backend='local_qasm_simulator_py',
shots=self.shots)
num_qubits = 2
feature_map = get_feature_map_instance('SecondOrderExpansion')
feature_map.init_args(num_qubits=num_qubits,
depth=2,
entangler_map={0: [1]})
svm.init_args(self.training_data, self.testing_data, None, feature_map,
None)
result = svm.run()
np.testing.assert_array_almost_equal(result['kernel_matrix_training'],
self.ref_kernel_matrix_training,
decimal=4)
np.testing.assert_array_almost_equal(result['kernel_matrix_testing'],
self.ref_kernel_matrix_testing,
decimal=4)
self.assertEqual(len(result['svm']['support_vectors']), 4)
np.testing.assert_array_almost_equal(result['svm']['support_vectors'],
self.ref_support_vectors,
decimal=4)
np.testing.assert_array_almost_equal(result['svm']['alphas'],
self.ref_alpha,
decimal=4)
np.testing.assert_array_almost_equal(result['svm']['bias'],
self.ref_bias,
decimal=4)
self.assertEqual(result['testing_accuracy'], 0.5)
def test_svm_variational_directly(self):
np.random.seed(self.random_seed)
svm = get_algorithm_instance("QSVM.Variational")
svm.random_seed = self.random_seed
svm.setup_quantum_backend(backend='local_qasm_simulator', shots=1024)
optimizer = get_optimizer_instance('SPSA')
optimizer.init_args(max_trials=10, c0=4.0, skip_calibration=True)
optimizer.set_options(save_steps=1)
num_qubits = 2
feature_map = get_feature_map_instance('SecondOrderExpansion')
feature_map.init_args(num_qubits=num_qubits, depth=2)
var_form = get_variational_form_instance('RYRZ')
var_form.init_args(num_qubits=num_qubits, depth=3)
svm.init_args(self.training_data, self.testing_data, None, optimizer,
feature_map, var_form)
result = svm.run()
np.testing.assert_array_almost_equal(result['opt_params'],
self.ref_opt_params,
decimal=4)
np.testing.assert_array_almost_equal(result['training_loss'],
self.ref_train_loss,
decimal=8)
self.assertEqual(result['testing_accuracy'], 0.5)
def init_params(self, params, algo_input):
fea_map_params = params.get(QuantumAlgorithm.SECTION_KEY_FEATURE_MAP)
feature_map = get_feature_map_instance(fea_map_params['name'])
num_qubits = get_feature_dimension(algo_input.training_dataset)
fea_map_params['num_qubits'] = num_qubits
feature_map.init_params(fea_map_params)
multiclass_extension = None
multiclass_extension_params = params.get(QuantumAlgorithm.SECTION_KEY_MULTICLASS_EXTENSION)
if multiclass_extension_params is not None:
multiclass_extension = get_multiclass_extension_instance(multiclass_extension_params['name'])
multiclass_extension_params['params'] = [feature_map, self]
multiclass_extension_params['estimator_cls'] = QSVM_Kernel_Estimator
multiclass_extension.init_params(multiclass_extension_params)
logger.info("Multiclass dataset with extension: {}".format(multiclass_extension_params['name']))
self.init_args(algo_input.training_dataset, algo_input.test_dataset,
algo_input.datapoints, feature_map, multiclass_extension)
sample_Total, training_input, test_input, class_labels = datasets.userDefinedData(
file='bank1000.csv',
location='',
training_size=training_dataset_size,
test_size=testing_dataset_size,
class_labels=[0, 1],
n=n,
PLOT_DATA=False)
datapoints, class_to_label = split_dataset_to_data_and_labels(test_input)
svm = get_algorithm_instance("QSVM.Kernel")
svm.random_seed = 10598
svm.setup_quantum_backend(backend='statevector_simulator')
feature_map = get_feature_map_instance('SecondOrderExpansion')
feature_map.init_args(num_qubits=2, depth=2, entanglement='linear')
svm.init_args(training_input, test_input, datapoints[0], feature_map)
print("Running algo")
result = svm.run()
print("kernel matrix during the training:")
kernel_matrix = result['kernel_matrix_training']
img = plt.imshow(np.asmatrix(kernel_matrix),
interpolation='nearest',
origin='upper',
cmap='bone_r')
plt.show()