A test case specifies an input scenario and expected result for evaluating software behavior. In quantum programming, automated tests can run circuits on a simulator and compare measured or internal states with expectations. section 2.4
Effective test cases expose distinct properties of the problem rather than checking only an easy success case. State-preparation tests should include varied amplitudes, negative values, zero values, basis states, and cases that reveal qubit-ordering errors. section 2.4
A quantum test case can validate behavior by:
- Preparing a chosen basis state and measuring it repeatedly; deterministic behavior should produce one result for every shot. section 4.1.2
- Converting between integer basis-state values and bit strings while keeping little-endian and big-endian conventions consistent. section 4.1.2
- Running the same validation over randomly selected qubit counts and basis states to broaden coverage. section 4.1.2
- Comparing reconstructed state coefficients with the actual coefficients using a classical closeness check. section 4.5.2
Testing a basis-state readout
A Qiskit helper accepts a qubit count and integer basis state, applies X gates for set bits, measures 100 shots, asserts that the result map has one key, and checks that the decoded bit string equals the input. section 4.1.2
Listing 4.2 Qiskit code validating the information readout
from random import randint
from qiskit import QuantumCircuit, transpile
from qiskit_aer import AerSimulator
from .read_info import read_info
sim = AerSimulator(method='statevector')
def run_test_read_info(n, basis_state):
circ = QuantumCircuit(n, n)
for i in range(n): #1
if basis_state & (1 {\fontsize{9}{0}\selectfont<\hskip-1pt<} i) > 0: #2
circ.x(i)
circ.append(read_info(n), range(n), range(n)) #3
circ = transpile(circ, backend=simulator) #4
res_map = simulator.run(circ, shots=100).result().get_counts()
assert len(res_map) == 1 #5
res_bitstring = list(res_map.keys())[0]
res = int(res_bitstring, 2) #6
assert res == basis_state
def test_read_info():
for _ in range(1, 20): #7
n = randint(1, 5)
num = randint(0, 2 ** n - 1)
run_test_read_info(n, num)
Inspect the complete Qiskit test workflow, including basis-state preparation, simulation, shot aggregation, result decoding, assertions, and randomized repetition. section 4.1.2
