class TestLocalQasmSimulatorCpp(QiskitTestCase):
"""
Test job_processor module.
"""
def setUp(self):
self.seed = 88
self.qasm_filename = self._get_resource_path('qasm/example.qasm')
with open(self.qasm_filename, 'r') as qasm_file:
self.qasm_text = qasm_file.read()
self.qasm_ast = qiskit.qasm.Qasm(data=self.qasm_text).parse()
self.qasm_be = qiskit.unroll.CircuitBackend(['u1', 'u2', 'u3', 'id', 'cx'])
self.qasm_circ = qiskit.unroll.Unroller(self.qasm_ast, self.qasm_be).execute()
qr = QuantumRegister(2, 'q')
cr = ClassicalRegister(2, 'c')
qc = QuantumCircuit(qr, cr)
qc.h(qr[0])
qc.measure(qr[0], cr[0])
self.qc = qc
# create qobj
compiled_circuit1 = qiskit._compiler.compile_circuit(self.qc, format='json')
compiled_circuit2 = qiskit._compiler.compile_circuit(self.qasm_circ, format='json')
self.qobj = {'id': 'test_qobj',
'config': {
'max_credits': 3,
'shots': 2000,
'backend_name': 'local_qasm_simulator_cpp',
'seed': 1111
},
'circuits': [
{
'name': 'test_circuit1',
'compiled_circuit': compiled_circuit1,
'basis_gates': 'u1,u2,u3,cx,id',
'layout': None,
},
{
'name': 'test_circuit2',
'compiled_circuit': compiled_circuit2,
'basis_gates': 'u1,u2,u3,cx,id',
'layout': None,
}
]}
# Simulator backend
try:
self.backend = QasmSimulatorCpp()
except FileNotFoundError as fnferr:
raise unittest.SkipTest(
'cannot find {} in path'.format(fnferr))
self.q_job = QuantumJob(self.qobj,
backend=self.backend,
preformatted=True)
def test_x90_coherent_error_matrix(self):
X90 = np.array([[1, -1j], [-1j, 1]]) / np.sqrt(2)
U = x90_error_matrix(0., 0.).dot(X90)
target = X90
self.assertAlmostEqual(norm(U - target), 0.0, places=10,
msg="identity error matrix")
U = x90_error_matrix(np.pi / 2., 0.).dot(X90)
target = -1j * np.array([[0, 1], [1, 0]])
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = x90_error_matrix(0., np.pi / 2.).dot(X90)
target = np.array([[1., -1], [1, 1.]]) / np.sqrt(2.)
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = x90_error_matrix(np.pi / 2, np.pi / 2.).dot(X90)
target = np.array([[0., -1], [1, 0.]])
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = x90_error_matrix(0.02, -0.03)
self.assertAlmostEqual(norm(U.dot(U.conj().T) - np.eye(2)), 0.0,
places=10, msg="Test error matrix is unitary")
def test_cx_coherent_error_matrix(self):
CX = np.array([[1, 0, 0, 0], [0, 0, 0, 1], [0, 0, 1, 0], [0, 1, 0, 0]])
U = cx_error_matrix(0., 0.).dot(CX)
target = CX
self.assertAlmostEqual(norm(U - target), 0.0, places=10,
msg="identity error matrix")
U = cx_error_matrix(np.pi / 2., 0.).dot(CX)
target = np.array([[1, 0, 1j, 0],
[0, -1j, 0, 1],
[1j, 0, 1, 0],
[0, 1, 0, -1j]]) / np.sqrt(2)
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = cx_error_matrix(0.03, -0.04)
self.assertAlmostEqual(norm(U.dot(U.conj().T) - np.eye(4)), 0.0,
places=10, msg="Test error matrix is unitary")
def test_run_qobj(self):
result = self.backend.run(self.q_job).result()
shots = self.qobj['config']['shots']
threshold = 0.04 * shots
counts = result.get_counts('test_circuit2')
target = {'100 100': shots / 8, '011 011': shots / 8,
'101 101': shots / 8, '111 111': shots / 8,
'000 000': shots / 8, '010 010': shots / 8,
'110 110': shots / 8, '001 001': shots / 8}
self.assertDictAlmostEqual(counts, target, threshold)
def test_qobj_measure_opt(self):
filename = self._get_resource_path('qobj/cpp_measure_opt.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
shots = q_job.qobj['config']['shots']
expected_data = {
'measure (opt)': {
'deterministic': True,
'counts': {'00': shots},
'statevector': np.array([1, 0, 0, 0])},
'x0 measure (opt)': {
'deterministic': True,
'counts': {'01': shots},
'statevector': np.array([0, 1, 0, 0])},
'x1 measure (opt)': {
'deterministic': True,
'counts': {'10': shots},
'statevector': np.array([0, 0, 1, 0])},
'x0 x1 measure (opt)': {
'deterministic': True,
'counts': {'11': shots},
'statevector': np.array([0, 0, 0, 1])},
'y0 measure (opt)': {
'deterministic': True,
'counts': {'01': shots},
'statevector': np.array([0, 1j, 0, 0])},
'y1 measure (opt)': {
'deterministic': True,
'counts': {'10': shots},
'statevector': np.array([0, 0, 1j, 0])},
'y0 y1 measure (opt)': {
'deterministic': True,
'counts': {'11': shots},
'statevector': np.array([0, 0, 0, -1j])},
'h0 measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 2, '01': shots / 2},
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 2, '10': shots / 2},
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h0 h1 measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 4, '01': shots / 4,
'10': shots / 4, '11': shots / 4},
'statevector': np.array([0.5, 0.5, 0.5, 0.5])},
'bell measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 2, '11': shots / 2},
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])}
}
for name in expected_data:
# Check counts:
counts = result.get_counts(name)
expected_counts = expected_data[name]['counts']
if expected_data[name].get('deterministic', False):
self.assertEqual(counts, expected_counts,
msg=name + ' counts')
else:
threshold = 0.04 * shots
self.assertDictAlmostEqual(counts, expected_counts,
threshold, msg=name + 'counts')
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_qobj_measure_opt_flag(self):
filename = self._get_resource_path('qobj/cpp_measure_opt_flag.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
shots = q_job.qobj['config']['shots']
sampled_measurements = {
'measure (sampled)': True,
'trivial (sampled)': True,
'reset1 (shots)': False,
'reset2 (shots)': False,
'reset3 (shots)': False,
'gate1 (shots)': False,
'gate2 (shots)': False,
'gate3 (shots)': False,
'gate4 (shots)': False
}
for name in sampled_measurements:
snapshots = result.get_snapshots(name)
# Check snapshot keys
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
# Check number of snapshots
# there should be 1 for measurement sampling optimization
# and there should be >1 for each shot beign simulated.
num_snapshots = len(snapshots['0'].get('statevector', []))
if sampled_measurements[name] is True:
self.assertEqual(num_snapshots, 1,
msg=name + ' snapshot length')
else:
self.assertEqual(num_snapshots, shots,
msg=name + ' snapshot length')
def test_qobj_reset(self):
filename = self._get_resource_path('qobj/cpp_reset.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'reset': {'statevector': np.array([1, 0])},
'x reset': {'statevector': np.array([1, 0])},
'y reset': {'statevector': np.array([1, 0])},
'h reset': {'statevector': np.array([1, 0])}
}
for name in expected_data:
# Check snapshot is |0> state
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_qobj_save_load(self):
filename = self._get_resource_path('qobj/cpp_save_load.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
snapshots = result.get_snapshots('save_command')
self.assertEqual(set(snapshots), {'0', '1', '10', '11'},
msg='snapshot keys')
state0 = snapshots['0']['statevector'][0]
state10 = snapshots['10']['statevector'][0]
state1 = snapshots['1']['statevector'][0]
state11 = snapshots['11']['statevector'][0]
expected_state0 = np.array([1, 0])
expected_state10 = np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
fidelity0 = np.abs(expected_state0.dot(state0.conj())) ** 2
fidelity1 = np.abs(expected_state0.dot(state1.conj())) ** 2
fidelity10 = np.abs(expected_state10.dot(state10.conj())) ** 2
fidelity11 = np.abs(expected_state10.dot(state11.conj())) ** 2
self.assertAlmostEqual(fidelity0, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity10, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity1, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity11, 1.0, places=10, msg='snapshot 0')
def test_qobj_single_qubit_gates(self):
filename = self._get_resource_path('qobj/cpp_single_qubit_gates.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'snapshot': {
'statevector': np.array([1, 0])},
'id(U)': {
'statevector': np.array([1, 0])},
'id(u3)': {
'statevector': np.array([1, 0])},
'id(u1)': {
'statevector': np.array([1, 0])},
'id(direct)': {
'statevector': np.array([1, 0])},
'x(U)': {
'statevector': np.array([0, 1])},
'x(u3)': {
'statevector': np.array([0, 1])},
'x(direct)': {
'statevector': np.array([0, 1])},
'y(U)': {
'statevector': np.array([0, 1j])},
'y(u3)': {
'statevector': np.array([0, 1j])},
'y(direct)': {
'statevector': np.array([0, 1j])},
'h(U)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(u3)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(u2)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(direct)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(direct) z(U)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) z(u3)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) z(u1)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) z(direct)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) s(U)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) s(u3)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) s(u1)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) s(direct)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) sdg(U)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) sdg(u3)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) sdg(u1)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) sdg(direct)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) t(U)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) t(u3)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) t(u1)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) t(direct)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) tdg(U)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])},
'h(direct) tdg(u3)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])},
'h(direct) tdg(u1)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])},
'h(direct) tdg(direct)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
inner_product = expected_state.dot(state.conj())
self.assertAlmostEqual(inner_product, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_qobj_two_qubit_gates(self):
filename = self._get_resource_path('qobj/cpp_two_qubit_gates.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'h0 CX01': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 CX10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 CX01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h1 CX10': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 cx01': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 cx10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 cx01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h1 cx10': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 cz01': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h0 cz10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 cz01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h1 cz10': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h0 h1 cz01': {'statevector': np.array([0.5, 0.5, 0.5, -0.5])},
'h0 h1 cz10': {'statevector': np.array([0.5, 0.5, 0.5, -0.5])},
'h0 rzz01': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2), 0, 0])},
'h0 rzz10': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2), 0, 0])},
'h1 rzz01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1j / np.sqrt(2), 0])},
'h1 rzz10': {
'statevector': np.array([1 / np.sqrt(2), 0, 1j / np.sqrt(2), 0])},
'h0 h1 rzz01': {'statevector': np.array([0.5, 0.5j, 0.5j, 0.5])},
'h0 h1 rzz10': {'statevector': np.array([0.5, 0.5j, 0.5j, 0.5])}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_conditionals(self):
filename = self._get_resource_path('qobj/cpp_conditionals.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'single creg (c0=0)': {
'statevector': np.array([1, 0, 0, 0])},
'single creg (c0=1)': {
'statevector': np.array([0, 0, 0, 1])},
'two creg (c1=0)': {
'statevector': np.array([1, 0, 0, 0])},
'two creg (c1=1)': {
'statevector': np.array([0, 0, 0, 1])}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
class TestLocalQasmSimulatorCpp(QiskitTestCase):
"""
Test job_processor module.
"""
def setUp(self):
self.seed = 88
self.qasm_filename = self._get_resource_path('qasm/example.qasm')
with open(self.qasm_filename, 'r') as qasm_file:
self.qasm_text = qasm_file.read()
self.qasm_ast = qiskit.qasm.Qasm(data=self.qasm_text).parse()
self.qasm_be = qiskit.unroll.CircuitBackend(
['u1', 'u2', 'u3', 'id', 'cx'])
self.qasm_circ = qiskit.unroll.Unroller(self.qasm_ast,
self.qasm_be).execute()
qr = QuantumRegister(2, 'q')
cr = ClassicalRegister(2, 'c')
qc = QuantumCircuit(qr, cr)
qc.h(qr[0])
qc.measure(qr[0], cr[0])
self.qc = qc
# create qobj
compiled_circuit1 = qiskit._compiler.compile_circuit(self.qc,
format='json')
compiled_circuit2 = qiskit._compiler.compile_circuit(self.qasm_circ,
format='json')
self.qobj = {
'id':
'test_qobj',
'config': {
'max_credits': 3,
'shots': 2000,
'backend_name': 'local_qasm_simulator_cpp',
'seed': 1111
},
'circuits': [{
'name': 'test_circuit1',
'compiled_circuit': compiled_circuit1,
'basis_gates': 'u1,u2,u3,cx,id',
'layout': None,
}, {
'name': 'test_circuit2',
'compiled_circuit': compiled_circuit2,
'basis_gates': 'u1,u2,u3,cx,id',
'layout': None,
}]
}
# Simulator backend
try:
self.backend = QasmSimulatorCpp()
except FileNotFoundError as fnferr:
raise unittest.SkipTest('cannot find {} in path'.format(fnferr))
self.q_job = QuantumJob(self.qobj,
backend=self.backend,
preformatted=True)
def test_x90_coherent_error_matrix(self):
X90 = np.array([[1, -1j], [-1j, 1]]) / np.sqrt(2)
U = x90_error_matrix(0., 0.).dot(X90)
target = X90
self.assertAlmostEqual(norm(U - target),
0.0,
places=10,
msg="identity error matrix")
U = x90_error_matrix(np.pi / 2., 0.).dot(X90)
target = -1j * np.array([[0, 1], [1, 0]])
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = x90_error_matrix(0., np.pi / 2.).dot(X90)
target = np.array([[1., -1], [1, 1.]]) / np.sqrt(2.)
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = x90_error_matrix(np.pi / 2, np.pi / 2.).dot(X90)
target = np.array([[0., -1], [1, 0.]])
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = x90_error_matrix(0.02, -0.03)
self.assertAlmostEqual(norm(U.dot(U.conj().T) - np.eye(2)),
0.0,
places=10,
msg="Test error matrix is unitary")
def test_cx_coherent_error_matrix(self):
CX = np.array([[1, 0, 0, 0], [0, 0, 0, 1], [0, 0, 1, 0], [0, 1, 0, 0]])
U = cx_error_matrix(0., 0.).dot(CX)
target = CX
self.assertAlmostEqual(norm(U - target),
0.0,
places=10,
msg="identity error matrix")
U = cx_error_matrix(np.pi / 2., 0.).dot(CX)
target = np.array([[1, 0, 1j, 0], [0, -1j, 0, 1], [1j, 0, 1, 0],
[0, 1, 0, -1j]]) / np.sqrt(2)
self.assertAlmostEqual(norm(U - target), 0.0, places=10)
U = cx_error_matrix(0.03, -0.04)
self.assertAlmostEqual(norm(U.dot(U.conj().T) - np.eye(4)),
0.0,
places=10,
msg="Test error matrix is unitary")
def test_run_qobj(self):
result = self.backend.run(self.q_job).result()
shots = self.qobj['config']['shots']
threshold = 0.04 * shots
counts = result.get_counts('test_circuit2')
target = {
'100 100': shots / 8,
'011 011': shots / 8,
'101 101': shots / 8,
'111 111': shots / 8,
'000 000': shots / 8,
'010 010': shots / 8,
'110 110': shots / 8,
'001 001': shots / 8
}
self.assertDictAlmostEqual(counts, target, threshold)
def test_qobj_measure_opt(self):
filename = self._get_resource_path('qobj/cpp_measure_opt.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
shots = q_job.qobj['config']['shots']
expected_data = {
'measure (opt)': {
'deterministic': True,
'counts': {
'00': shots
},
'statevector': np.array([1, 0, 0, 0])
},
'x0 measure (opt)': {
'deterministic': True,
'counts': {
'01': shots
},
'statevector': np.array([0, 1, 0, 0])
},
'x1 measure (opt)': {
'deterministic': True,
'counts': {
'10': shots
},
'statevector': np.array([0, 0, 1, 0])
},
'x0 x1 measure (opt)': {
'deterministic': True,
'counts': {
'11': shots
},
'statevector': np.array([0, 0, 0, 1])
},
'y0 measure (opt)': {
'deterministic': True,
'counts': {
'01': shots
},
'statevector': np.array([0, 1j, 0, 0])
},
'y1 measure (opt)': {
'deterministic': True,
'counts': {
'10': shots
},
'statevector': np.array([0, 0, 1j, 0])
},
'y0 y1 measure (opt)': {
'deterministic': True,
'counts': {
'11': shots
},
'statevector': np.array([0, 0, 0, -1j])
},
'h0 measure (opt)': {
'deterministic': False,
'counts': {
'00': shots / 2,
'01': shots / 2
},
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])
},
'h1 measure (opt)': {
'deterministic': False,
'counts': {
'00': shots / 2,
'10': shots / 2
},
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])
},
'h0 h1 measure (opt)': {
'deterministic': False,
'counts': {
'00': shots / 4,
'01': shots / 4,
'10': shots / 4,
'11': shots / 4
},
'statevector': np.array([0.5, 0.5, 0.5, 0.5])
},
'bell measure (opt)': {
'deterministic': False,
'counts': {
'00': shots / 2,
'11': shots / 2
},
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])
}
}
for name in expected_data:
# Check counts:
counts = result.get_counts(name)
expected_counts = expected_data[name]['counts']
if expected_data[name].get('deterministic', False):
self.assertEqual(counts, expected_counts, msg=name + ' counts')
else:
threshold = 0.04 * shots
self.assertDictAlmostEqual(counts,
expected_counts,
threshold,
msg=name + 'counts')
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']),
3,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj()))**2
self.assertAlmostEqual(fidelity,
1.0,
places=10,
msg=name + ' snapshot fidelity')
rho = snapshots['0']['density_matrix']
self.assertAlmostEqual(np.trace(rho), 1)
prob = snapshots['0']['probabilities']
self.assertAlmostEqual(np.sum(prob), 1)
def test_qobj_measure_opt_flag(self):
filename = self._get_resource_path('qobj/cpp_measure_opt_flag.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
shots = q_job.qobj['config']['shots']
sampled_measurements = {
'measure (sampled)': True,
'trivial (sampled)': True,
'reset1 (shots)': False,
'reset2 (shots)': False,
'reset3 (shots)': False,
'gate1 (shots)': False,
'gate2 (shots)': False,
'gate3 (shots)': False,
'gate4 (shots)': False
}
for name in sampled_measurements:
snapshots = result.get_snapshots(name)
# Check snapshot keys
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
# Check number of snapshots
# there should be 1 for measurement sampling optimization
# and there should be >1 for each shot beign simulated.
num_snapshots = len(snapshots['0'].get('statevector', []))
if sampled_measurements[name] is True:
self.assertEqual(num_snapshots,
1,
msg=name + ' snapshot length')
else:
self.assertEqual(num_snapshots,
shots,
msg=name + ' snapshot length')
def test_qobj_reset(self):
filename = self._get_resource_path('qobj/cpp_reset.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'reset': {
'statevector': np.array([1, 0])
},
'x reset': {
'statevector': np.array([1, 0])
},
'y reset': {
'statevector': np.array([1, 0])
},
'h reset': {
'statevector': np.array([1, 0])
}
}
for name in expected_data:
# Check snapshot is |0> state
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']),
1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj()))**2
self.assertAlmostEqual(fidelity,
1.0,
places=10,
msg=name + ' snapshot fidelity')
def test_qobj_save_load(self):
filename = self._get_resource_path('qobj/cpp_save_load.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
snapshots = result.get_snapshots('save_command')
self.assertEqual(set(snapshots), {'0', '1', '10', '11'},
msg='snapshot keys')
state0 = snapshots['0']['statevector'][0]
state10 = snapshots['10']['statevector'][0]
state1 = snapshots['1']['statevector'][0]
state11 = snapshots['11']['statevector'][0]
expected_state0 = np.array([1, 0])
expected_state10 = np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
fidelity0 = np.abs(expected_state0.dot(state0.conj()))**2
fidelity1 = np.abs(expected_state0.dot(state1.conj()))**2
fidelity10 = np.abs(expected_state10.dot(state10.conj()))**2
fidelity11 = np.abs(expected_state10.dot(state11.conj()))**2
self.assertAlmostEqual(fidelity0, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity10, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity1, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity11, 1.0, places=10, msg='snapshot 0')
def test_qobj_single_qubit_gates(self):
filename = self._get_resource_path('qobj/cpp_single_qubit_gates.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'snapshot': {
'statevector': np.array([1, 0])
},
'id(U)': {
'statevector': np.array([1, 0])
},
'id(u3)': {
'statevector': np.array([1, 0])
},
'id(u1)': {
'statevector': np.array([1, 0])
},
'id(direct)': {
'statevector': np.array([1, 0])
},
'x(U)': {
'statevector': np.array([0, 1])
},
'x(u3)': {
'statevector': np.array([0, 1])
},
'x(direct)': {
'statevector': np.array([0, 1])
},
'y(U)': {
'statevector': np.array([0, 1j])
},
'y(u3)': {
'statevector': np.array([0, 1j])
},
'y(direct)': {
'statevector': np.array([0, 1j])
},
'h(U)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
},
'h(u3)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
},
'h(u2)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
},
'h(direct)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
},
'h(direct) z(U)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])
},
'h(direct) z(u3)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])
},
'h(direct) z(u1)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])
},
'h(direct) z(direct)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])
},
'h(direct) s(U)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])
},
'h(direct) s(u3)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])
},
'h(direct) s(u1)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])
},
'h(direct) s(direct)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])
},
'h(direct) sdg(U)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])
},
'h(direct) sdg(u3)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])
},
'h(direct) sdg(u1)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])
},
'h(direct) sdg(direct)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])
},
'h(direct) t(U)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])
},
'h(direct) t(u3)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])
},
'h(direct) t(u1)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])
},
'h(direct) t(direct)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])
},
'h(direct) tdg(U)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])
},
'h(direct) tdg(u3)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])
},
'h(direct) tdg(u1)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])
},
'h(direct) tdg(direct)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])
}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']),
1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
inner_product = expected_state.dot(state.conj())
self.assertAlmostEqual(inner_product,
1.0,
places=10,
msg=name + ' snapshot fidelity')
def test_qobj_two_qubit_gates(self):
filename = self._get_resource_path('qobj/cpp_two_qubit_gates.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'h0 CX01': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])
},
'h0 CX10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])
},
'h1 CX01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])
},
'h1 CX10': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])
},
'h0 cx01': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])
},
'h0 cx10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])
},
'h1 cx01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])
},
'h1 cx10': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])
},
'h0 cz01': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])
},
'h0 cz10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])
},
'h1 cz01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])
},
'h1 cz10': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])
},
'h0 h1 cz01': {
'statevector': np.array([0.5, 0.5, 0.5, -0.5])
},
'h0 h1 cz10': {
'statevector': np.array([0.5, 0.5, 0.5, -0.5])
},
'h0 rzz01': {
'statevector':
np.array([1 / np.sqrt(2), 1j / np.sqrt(2), 0, 0])
},
'h0 rzz10': {
'statevector':
np.array([1 / np.sqrt(2), 1j / np.sqrt(2), 0, 0])
},
'h1 rzz01': {
'statevector':
np.array([1 / np.sqrt(2), 0, 1j / np.sqrt(2), 0])
},
'h1 rzz10': {
'statevector':
np.array([1 / np.sqrt(2), 0, 1j / np.sqrt(2), 0])
},
'h0 h1 rzz01': {
'statevector': np.array([0.5, 0.5j, 0.5j, 0.5])
},
'h0 h1 rzz10': {
'statevector': np.array([0.5, 0.5j, 0.5j, 0.5])
}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']),
1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj()))**2
self.assertAlmostEqual(fidelity,
1.0,
places=10,
msg=name + ' snapshot fidelity')
def test_conditionals(self):
filename = self._get_resource_path('qobj/cpp_conditionals.json')
with open(filename, 'r') as file:
q_job = QuantumJob(json.load(file),
backend=self.backend,
preformatted=True)
result = self.backend.run(q_job).result()
expected_data = {
'single creg (c0=0)': {
'statevector': np.array([1, 0, 0, 0])
},
'single creg (c0=1)': {
'statevector': np.array([0, 0, 0, 1])
},
'two creg (c1=0)': {
'statevector': np.array([1, 0, 0, 0])
},
'two creg (c1=1)': {
'statevector': np.array([0, 0, 0, 1])
}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']),
1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj()))**2
self.assertAlmostEqual(fidelity,
1.0,
places=10,
msg=name + ' snapshot fidelity')
class TestLocalQasmSimulatorCpp(QiskitTestCase):
"""
Test job_processor module.
"""
@requires_cpp_simulator
def setUp(self):
self.seed = 88
self.qasm_filename = self._get_resource_path('qasm/example.qasm')
with open(self.qasm_filename, 'r') as qasm_file:
self.qasm_text = qasm_file.read()
self.qasm_ast = qiskit.qasm.Qasm(data=self.qasm_text).parse()
self.qasm_be = qiskit.unroll.CircuitBackend(['u1', 'u2', 'u3', 'id', 'cx'])
self.qasm_circ = qiskit.unroll.Unroller(self.qasm_ast, self.qasm_be).execute()
qr = QuantumRegister(2, 'q')
cr = ClassicalRegister(2, 'c')
qc = QuantumCircuit(qr, cr)
qc.h(qr[0])
qc.measure(qr[0], cr[0])
self.qc = qc
# create qobj
compiled_circuit1 = QobjExperiment.from_dict(
transpile(DAGCircuit.fromQuantumCircuit(self.qc), format='json'))
compiled_circuit2 = QobjExperiment.from_dict(
transpile(DAGCircuit.fromQuantumCircuit(self.qasm_circ),
format='json'))
self.qobj = Qobj(
qobj_id='test_qobj',
config=QobjConfig(
shots=2000, memory_slots=1,
max_credits=3,
seed=1111
),
experiments=[compiled_circuit1, compiled_circuit2],
header=QobjHeader(backend_name='local_qasm_simulator_cpp')
)
self.qobj.experiments[0].header.name = 'test_circuit1'
self.qobj.experiments[0].config = QobjItem(basis_gates='u1,u2,u3,cx,id')
self.qobj.experiments[1].header.name = 'test_circuit2'
self.qobj.experiments[1].config = QobjItem(basis_gates='u1,u2,u3,cx,id')
self.backend = QasmSimulatorCpp()
def test_x90_coherent_error_matrix(self):
x90 = np.array([[1, -1j], [-1j, 1]]) / np.sqrt(2)
u_matrix = x90_error_matrix(0., 0.).dot(x90)
target = x90
self.assertAlmostEqual(norm(u_matrix - target), 0.0, places=10,
msg="identity error matrix")
u_matrix = x90_error_matrix(np.pi / 2., 0.).dot(x90)
target = -1j * np.array([[0, 1], [1, 0]])
self.assertAlmostEqual(norm(u_matrix - target), 0.0, places=10)
u_matrix = x90_error_matrix(0., np.pi / 2.).dot(x90)
target = np.array([[1., -1], [1, 1.]]) / np.sqrt(2.)
self.assertAlmostEqual(norm(u_matrix - target), 0.0, places=10)
u_matrix = x90_error_matrix(np.pi / 2, np.pi / 2.).dot(x90)
target = np.array([[0., -1], [1, 0.]])
self.assertAlmostEqual(norm(u_matrix - target), 0.0, places=10)
u_matrix = x90_error_matrix(0.02, -0.03)
self.assertAlmostEqual(norm(u_matrix.dot(u_matrix.conj().T) - np.eye(2)), 0.0,
places=10, msg="Test error matrix is unitary")
def test_cx_coherent_error_matrix(self):
cx_matrix = np.array([[1, 0, 0, 0], [0, 0, 0, 1], [0, 0, 1, 0], [0, 1, 0, 0]])
u_matrix = cx_error_matrix(0., 0.).dot(cx_matrix)
target = cx_matrix
self.assertAlmostEqual(norm(u_matrix - target), 0.0, places=10,
msg="identity error matrix")
u_matrix = cx_error_matrix(np.pi / 2., 0.).dot(cx_matrix)
target = np.array([[1, 0, 1j, 0],
[0, -1j, 0, 1],
[1j, 0, 1, 0],
[0, 1, 0, -1j]]) / np.sqrt(2)
self.assertAlmostEqual(norm(u_matrix - target), 0.0, places=10)
u_matrix = cx_error_matrix(0.03, -0.04)
self.assertAlmostEqual(norm(u_matrix.dot(u_matrix.conj().T) - np.eye(4)), 0.0,
places=10, msg="Test error matrix is unitary")
def test_run_qobj(self):
result = self.backend.run(self.qobj).result()
shots = self.qobj.config.shots
threshold = 0.04 * shots
counts = result.get_counts('test_circuit2')
target = {'100 100': shots / 8, '011 011': shots / 8,
'101 101': shots / 8, '111 111': shots / 8,
'000 000': shots / 8, '010 010': shots / 8,
'110 110': shots / 8, '001 001': shots / 8}
self.assertDictAlmostEqual(counts, target, threshold)
def test_qobj_measure_opt(self):
filename = self._get_resource_path('qobj/cpp_measure_opt.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
shots = qobj.config.shots
expected_data = {
'measure (opt)': {
'deterministic': True,
'counts': {'00': shots},
'statevector': np.array([1, 0, 0, 0])},
'x0 measure (opt)': {
'deterministic': True,
'counts': {'01': shots},
'statevector': np.array([0, 1, 0, 0])},
'x1 measure (opt)': {
'deterministic': True,
'counts': {'10': shots},
'statevector': np.array([0, 0, 1, 0])},
'x0 x1 measure (opt)': {
'deterministic': True,
'counts': {'11': shots},
'statevector': np.array([0, 0, 0, 1])},
'y0 measure (opt)': {
'deterministic': True,
'counts': {'01': shots},
'statevector': np.array([0, 1j, 0, 0])},
'y1 measure (opt)': {
'deterministic': True,
'counts': {'10': shots},
'statevector': np.array([0, 0, 1j, 0])},
'y0 y1 measure (opt)': {
'deterministic': True,
'counts': {'11': shots},
'statevector': np.array([0, 0, 0, -1j])},
'h0 measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 2, '01': shots / 2},
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 2, '10': shots / 2},
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h0 h1 measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 4, '01': shots / 4,
'10': shots / 4, '11': shots / 4},
'statevector': np.array([0.5, 0.5, 0.5, 0.5])},
'bell measure (opt)': {
'deterministic': False,
'counts': {'00': shots / 2, '11': shots / 2},
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])}
}
for name in expected_data:
# Check counts:
counts = result.get_counts(name)
expected_counts = expected_data[name]['counts']
if expected_data[name].get('deterministic', False):
self.assertEqual(counts, expected_counts,
msg=name + ' counts')
else:
threshold = 0.04 * shots
self.assertDictAlmostEqual(counts, expected_counts,
threshold, msg=name + 'counts')
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 3,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
rho = snapshots['0']['density_matrix']
self.assertAlmostEqual(np.trace(rho), 1)
prob = snapshots['0']['probabilities']
self.assertAlmostEqual(np.sum(prob), 1)
def test_qobj_measure_opt_flag(self):
filename = self._get_resource_path('qobj/cpp_measure_opt_flag.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
shots = qobj.config.shots
sampled_measurements = {
'measure (sampled)': True,
'trivial (sampled)': True,
'reset1 (shots)': False,
'reset2 (shots)': False,
'reset3 (shots)': False,
'gate1 (shots)': False,
'gate2 (shots)': False,
'gate3 (shots)': False,
'gate4 (shots)': False
}
for name in sampled_measurements:
snapshots = result.get_snapshots(name)
# Check snapshot keys
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
# Check number of snapshots
# there should be 1 for measurement sampling optimization
# and there should be >1 for each shot beign simulated.
num_snapshots = len(snapshots['0'].get('statevector', []))
if sampled_measurements[name] is True:
self.assertEqual(num_snapshots, 1,
msg=name + ' snapshot length')
else:
self.assertEqual(num_snapshots, shots,
msg=name + ' snapshot length')
def test_qobj_reset(self):
filename = self._get_resource_path('qobj/cpp_reset.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
expected_data = {
'reset': {'statevector': np.array([1, 0])},
'x reset': {'statevector': np.array([1, 0])},
'y reset': {'statevector': np.array([1, 0])},
'h reset': {'statevector': np.array([1, 0])}
}
for name in expected_data:
# Check snapshot is |0> state
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_qobj_save_load(self):
filename = self._get_resource_path('qobj/cpp_save_load.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
snapshots = result.get_snapshots('save_command')
self.assertEqual(set(snapshots), {'0', '1', '10', '11'},
msg='snapshot keys')
state0 = snapshots['0']['statevector'][0]
state10 = snapshots['10']['statevector'][0]
state1 = snapshots['1']['statevector'][0]
state11 = snapshots['11']['statevector'][0]
expected_state0 = np.array([1, 0])
expected_state10 = np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])
fidelity0 = np.abs(expected_state0.dot(state0.conj())) ** 2
fidelity1 = np.abs(expected_state0.dot(state1.conj())) ** 2
fidelity10 = np.abs(expected_state10.dot(state10.conj())) ** 2
fidelity11 = np.abs(expected_state10.dot(state11.conj())) ** 2
self.assertAlmostEqual(fidelity0, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity10, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity1, 1.0, places=10, msg='snapshot 0')
self.assertAlmostEqual(fidelity11, 1.0, places=10, msg='snapshot 0')
def test_qobj_single_qubit_gates(self):
filename = self._get_resource_path('qobj/cpp_single_qubit_gates.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
expected_data = {
'snapshot': {
'statevector': np.array([1, 0])},
'id(U)': {
'statevector': np.array([1, 0])},
'id(u3)': {
'statevector': np.array([1, 0])},
'id(u1)': {
'statevector': np.array([1, 0])},
'id(direct)': {
'statevector': np.array([1, 0])},
'x(U)': {
'statevector': np.array([0, 1])},
'x(u3)': {
'statevector': np.array([0, 1])},
'x(direct)': {
'statevector': np.array([0, 1])},
'y(U)': {
'statevector': np.array([0, 1j])},
'y(u3)': {
'statevector': np.array([0, 1j])},
'y(direct)': {
'statevector': np.array([0, 1j])},
'h(U)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(u3)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(u2)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(direct)': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2)])},
'h(direct) z(U)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) z(u3)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) z(u1)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) z(direct)': {
'statevector': np.array([1 / np.sqrt(2), -1 / np.sqrt(2)])},
'h(direct) s(U)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) s(u3)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) s(u1)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) s(direct)': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2)])},
'h(direct) sdg(U)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) sdg(u3)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) sdg(u1)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) sdg(direct)': {
'statevector': np.array([1 / np.sqrt(2), -1j / np.sqrt(2)])},
'h(direct) t(U)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) t(u3)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) t(u1)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) t(direct)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 + 0.5j])},
'h(direct) tdg(U)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])},
'h(direct) tdg(u3)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])},
'h(direct) tdg(u1)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])},
'h(direct) tdg(direct)': {
'statevector': np.array([1 / np.sqrt(2), 0.5 - 0.5j])}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
inner_product = expected_state.dot(state.conj())
self.assertAlmostEqual(inner_product, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_qobj_two_qubit_gates(self):
filename = self._get_resource_path('qobj/cpp_two_qubit_gates.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
expected_data = {
'h0 CX01': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 CX10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 CX01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h1 CX10': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 cx01': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 cx10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 cx01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h1 cx10': {
'statevector': np.array([1 / np.sqrt(2), 0, 0, 1 / np.sqrt(2)])},
'h0 cz01': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h0 cz10': {
'statevector': np.array([1 / np.sqrt(2), 1 / np.sqrt(2), 0, 0])},
'h1 cz01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h1 cz10': {
'statevector': np.array([1 / np.sqrt(2), 0, 1 / np.sqrt(2), 0])},
'h0 h1 cz01': {'statevector': np.array([0.5, 0.5, 0.5, -0.5])},
'h0 h1 cz10': {'statevector': np.array([0.5, 0.5, 0.5, -0.5])},
'h0 rzz01': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2), 0, 0])},
'h0 rzz10': {
'statevector': np.array([1 / np.sqrt(2), 1j / np.sqrt(2), 0, 0])},
'h1 rzz01': {
'statevector': np.array([1 / np.sqrt(2), 0, 1j / np.sqrt(2), 0])},
'h1 rzz10': {
'statevector': np.array([1 / np.sqrt(2), 0, 1j / np.sqrt(2), 0])},
'h0 h1 rzz01': {'statevector': np.array([0.5, 0.5j, 0.5j, 0.5])},
'h0 h1 rzz10': {'statevector': np.array([0.5, 0.5j, 0.5j, 0.5])}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
def test_conditionals(self):
filename = self._get_resource_path('qobj/cpp_conditionals.json')
with open(filename, 'r') as file:
qobj = Qobj.from_dict(json.load(file))
result = self.backend.run(qobj).result()
expected_data = {
'single creg (c0=0)': {
'statevector': np.array([1, 0, 0, 0])},
'single creg (c0=1)': {
'statevector': np.array([0, 0, 0, 1])},
'two creg (c1=0)': {
'statevector': np.array([1, 0, 0, 0])},
'two creg (c1=1)': {
'statevector': np.array([0, 0, 0, 1])}
}
for name in expected_data:
# Check snapshot
snapshots = result.get_snapshots(name)
self.assertEqual(set(snapshots), {'0'},
msg=name + ' snapshot keys')
self.assertEqual(len(snapshots['0']), 1,
msg=name + ' snapshot length')
state = snapshots['0']['statevector'][0]
expected_state = expected_data[name]['statevector']
fidelity = np.abs(expected_state.dot(state.conj())) ** 2
self.assertAlmostEqual(fidelity, 1.0, places=10,
msg=name + ' snapshot fidelity')
