pytest is a Python testing framework for discovering, organizing, and running automated tests. It finds functions beginning with test_ in files named test_*.py or *_test.py, keeping tests separate from library logic. section 2.4.1
Quantum programs can expose simulator state as ordinary Python data, allowing pytest assertions to compare computed amplitudes with expected values. Q# tests divide the test harness between Python and Q# project files. section 2.4.1 section 2.4.2
A typical quantum-state test workflow:
- Decorate a test with pytest.mark.parametrize to run it against several amplitude lists. section 2.4.1 section 2.4.2
- Run the Qiskit circuit on an AerSimulator statevector backend, retrieve its state vector, and compare it with pytest.approx. section 2.4.1
- For Q#, initialize the project with qsharp.init, evaluate the operation with qsharp.eval, retrieve the dense state, and assert its amplitudes. section 2.4.2
- Validate deterministic measurement by asserting one result key and converting its bit string to an integer. section 4.1.2
Testing two-qubit state preparation
A parametrized test prepares several two-qubit states, simulates the circuit, extracts the state vector, and checks state_vector == pytest.approx(a). section 2.4.1
Listing 2.7 Qiskit code to test preparation of an arbitrary state
from math import sqrt
from qiskit import transpile
from qiskit_aer import AerSimulator
import pytest
from .prep_two_qubit import prep_two_qubit
simulator = AerSimulator(method='statevector')
@pytest.mark.parametrize("a",
[ [1., 0., 0., 0.], #1
[0., 1., 0., 0.],
[0.36, 0.48, 0.64, -0.48],
[1. / sqrt(3.), -1. / sqrt(3.), 1. / sqrt(3.), 0.]
])
def test_prep_two_qubit(a):
circ = prep_two_qubit(a)
circ.save_statevector()
circ = transpile(circ, backend=simulator)
res = simulator.run(circ).result()
state_vector = res.get_statevector().data
assert state_vector == pytest.approx(a) #2
Inspect a complete Qiskit pytest listing using AerSimulator, parametrization, state-vector extraction, and approximate comparison. section 2.4.1
Listing 2.8 Python code to test Q# preparation of a two-qubit state
from math import sqrt
import qsharp
import pytest
@pytest.mark.parametrize("a",
[ [1., 0., 0., 0.],
[0., 1., 0., 0.], #1
[0.36, 0.48, 0.64, -0.48],
[1. / sqrt(3.), -1. / sqrt(3.), 1. / sqrt(3.), 0.]
])
def test_prep_two_qubit(a):
qsharp.init(project_root='.')
qsharp.eval(f"use qs = Qubit[2]; StatePreparation.PrepTwoQubits(qs, {a});")
state_vector = qsharp.dump_machine().as_dense_state()
assert state_vector == pytest.approx(a) #2
Inspect the corresponding Q# test harness, which loads the project, evaluates Q# code, dumps the simulated state, and validates amplitudes. section 2.4.2
