def test_file_error_model_generate():
fem = FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl'))
fqc = FiveQubitCode()
for _ in range(10):
error = fem.generate(fqc, 0.4)
assert isinstance(error, np.ndarray)
assert len(error) == 10
def test_file_error_model_generate_skip_to_start():
print()
fem = FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl'), 4)
fqc = FiveQubitCode()
packed_errors = (
pt.pack(fem.generate(fqc, 0.4)),
pt.pack(fem.generate(fqc, 0.4)),
pt.pack(fem.generate(fqc, 0.4)),
)
expected_packed_errors = (("8400", 10), ("5280", 10), ("1080", 10))
assert packed_errors == expected_packed_errors
def test_file_error_model_generate_no_more_errors_available():
fem = FileErrorModel(os.path.join(FILES_DIR, 'no_more_errors_available.jsonl'))
fqc = FiveQubitCode()
with pytest.raises(EOFError):
fem.generate(fqc, 0.4)
fem.generate(fqc, 0.4)
fem.generate(fqc, 0.4)
def test_file_error_model_generated_sample_error_probability(code, filename, decoder):
with CliRunner().isolated_filesystem(): # isolate from logging_qecsim.ini
# error model and probability from sample
error_model = FileErrorModel(filename, start=1000)
e_prob = error_model._probability
# runs (repeat many times to ensure physical_error_rate is close to error_probability)
max_runs = 100
data = app.run(code, error_model, decoder, e_prob, max_runs) # no error raised
p_rate = data['physical_error_rate']
p_var = data['error_weight_pvar'] / (data['n_k_d'][0] ** 2) # physical_error_rate_pvar (power of 2 is correct)
p_std = math.sqrt(p_var) # physical_error_rate_std
assert p_rate - p_std < e_prob < p_rate + p_std, (
'physical_error_rate={} is not within 1 std={} of error_probability={}'.format(p_rate, p_std, e_prob))
def test_file_error_model_probability_distribution():
fem = FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl'))
pd = fem.probability_distribution(fem._probability)
assert isinstance(pd, tuple)
assert len(pd) == 4
assert pd[0] + pd[1] + pd[2] + pd[3] == 1
def test_file_error_model_init_invalid_extra_headers(filename):
with pytest.raises(ValueError):
FileErrorModel(filename)
def test_file_error_model_init_extra_header():
fem = FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl'))
assert fem.bias == 10
def test_file_error_model_init_default_parameters():
FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl')) # no error raised
def test_file_error_model_file_not_found(filename):
with pytest.raises(FileNotFoundError):
FileErrorModel(filename)
def test_file_error_model_init_invalid_parameters(filename, start):
with pytest.raises((ValueError, TypeError), match=r"^FileErrorModel") as exc_info:
FileErrorModel(filename, start) # expected error raised
print(exc_info)
def test_file_error_model_init(filename, start):
FileErrorModel(filename, start) # no error raised
def test_file_error_model_generate_incorrect_length_in_packed_error(filename):
fem = FileErrorModel(filename)
fqc = FiveQubitCode()
with pytest.raises(ValueError):
fem.generate(fqc, 0.4)
def test_file_error_model_generate_invalid_error_lines():
fem = FileErrorModel(os.path.join(FILES_DIR, 'header_lines_after_body_lines.jsonl'))
fqc = FiveQubitCode()
with pytest.raises(ValueError):
fem.generate(fqc, 0.4)
fem.generate(fqc, 0.4)
def test_file_error_model_generate_invalid_probability_parameter():
fem = FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl'))
fqc = FiveQubitCode()
with pytest.raises(ValueError):
fem.generate(fqc, 0.3)
def test_file_error_model__probability_distribution_no_probability_distribution_in_header():
fem = FileErrorModel(os.path.join(FILES_DIR, 'no_probability_distribution.jsonl'))
with pytest.raises(ValueError):
fem.probability_distribution(0.4)
def test_file_error_model_probability_distribution_invalid_probability_parameter():
fem = FileErrorModel(os.path.join(FILES_DIR, 'fivequbitcode-errors-p0.4-bias10.jsonl'))
with pytest.raises(ValueError):
fem.probability_distribution(0.3)