def test_distinct_seeds():
rrs = samplespace.RepeatableRandomSequence(seed='abcdef')
run1 = [rrs.getnextblock() for _ in range(256)]
rrs = samplespace.RepeatableRandomSequence(seed=98471)
run2 = [rrs.getnextblock() for _ in range(256)]
assert run1 != run2
def test_no_cascading():
rrs = samplespace.RepeatableRandomSequence(seed=1234)
# Ensure no-cascade methods are blocked
with pytest.raises(RuntimeError):
with rrs.cascade():
rrs.seed(10)
with pytest.raises(RuntimeError):
with rrs.cascade():
rrs.index = 10
with pytest.raises(RuntimeError):
with rrs.cascade():
_ = rrs.index
with pytest.raises(RuntimeError):
with rrs.cascade():
rrs.reset()
with pytest.raises(RuntimeError):
with rrs.cascade():
rrs.getstate()
with pytest.raises(RuntimeError):
with rrs.cascade():
# noinspection PyTypeChecker
rrs.setstate(None)
def test_geometric_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.geometric(0.5, False)
with pytest.raises(ValueError):
rrs.geometric(-0.5, True)
def test_randint():
rrs = samplespace.RepeatableRandomSequence(seed='hello')
expected = [rrs.randrange(2, 14) for _ in range(10)]
rrs.reset()
actual = [rrs.randint(2, 13) for _ in range(10)]
assert actual == expected
def test_randbelow_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(TypeError):
rrs._randbelow(2.1)
with pytest.raises(ValueError):
rrs._randbelow(0)
def test_reset():
rrs = samplespace.RepeatableRandomSequence(seed='abcdef')
run1 = [rrs.getnextblock() for _ in range(256)]
rrs.reset()
run2 = [rrs.getnextblock() for _ in range(256)]
assert run1 == run2
def test_state():
# Restoring state
rrs = samplespace.RepeatableRandomSequence(seed=12345)
for _ in range(100):
rrs.getnextblock()
state = rrs.getstate()
expected = rrs.getnextblock()
for _ in range(100):
rrs.getnextblock()
rrs.setstate(state)
actual = rrs.getnextblock()
assert actual == expected
# Restoring state using indices
start_index = rrs.index
expected = rrs.getnextblock()
for _ in range(100):
rrs.getnextblock()
rrs.index = start_index
actual = rrs.getnextblock()
assert actual == expected
# Rewinding indices
expected = [rrs.random() for _ in range(10)]
for _ in range(50):
rrs.getnextblock()
rrs.index -= 60
actual = [rrs.random() for _ in range(10)]
assert actual == expected
# Copying state
state = rrs.getstate()
expected = rrs.getnextblock()
rrs2 = samplespace.RepeatableRandomSequence(seed='hello')
rrs2.setstate(state)
actual = rrs2.getnextblock()
assert actual == expected
def test_sample_args():
rrs = samplespace.RepeatableRandomSequence()
assert rrs.sample([1], 0) == []
with pytest.raises(IndexError):
rrs.sample([], 1)
with pytest.raises(ValueError):
rrs.sample([1, 2, 3], 5)
def test_triangular_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.triangular(2.0, 1.0)
with pytest.raises(ValueError):
rrs.triangular(1.0, 2.0, 3.0)
assert rrs.triangular(2.0, 2.0, 2.0) == 2.0
def test_choices_args():
rrs = samplespace.RepeatableRandomSequence()
assert rrs.choices([1, 2], k=0) == []
with pytest.raises(IndexError):
rrs.choices([], k=1)
with pytest.raises(TypeError):
rrs.choices([1], weights=[1.0], cum_weights=[1.0])
with pytest.raises(ValueError):
rrs.choices([1], weights=[1.0, 2.0])
def test_shuffle():
rrs = samplespace.RepeatableRandomSequence(seed='hello')
initial = list(range(100))
start_index = rrs.index
shuffled = list(initial)
rrs.shuffle(shuffled)
assert set(initial) == set(shuffled)
assert initial != shuffled
assert rrs.index == start_index + 1
def test_randrange_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(TypeError):
rrs.randrange(2.1)
with pytest.raises(TypeError):
rrs.randrange(1, 2.1)
with pytest.raises(TypeError):
rrs.randrange(1, 2, 0.5)
with pytest.raises(ValueError):
rrs.randrange(1, 2, 0)
def test_serialize_rrs_state():
"""Verifies that RRS states serialize to YAML correctly."""
rrs = samplespace.RepeatableRandomSequence(seed=12345)
[rrs.getnextblock() for _ in range(100)]
state = rrs.getstate()
expected = rrs.getnextblock()
state_as_dict = state.as_dict()
[rrs.getnextblock() for _ in range(100)]
new_state = samplespace.repeatablerandom.RepeatableRandomSequenceState.from_dict(
state_as_dict)
assert new_state == state
rrs.setstate(new_state)
actual = rrs.getnextblock()
assert actual == expected
def test_zipf_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.zipfmandelbrot(1.5, 1.0, 0)
def test_weibull_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.weibullvariate(1.0, 0.0)
def test_expected_sequence():
# Test raw 64 bit chunks for seed 0
rrs = samplespace.RepeatableRandomSequence(seed=0)
actual = [rrs.getnextblock() for _ in range(10)]
expected = test_data['raw-seed0-index0-n10']
assert actual == expected
# Test raw 64 bit chunks for a different seed
rrs = samplespace.RepeatableRandomSequence(seed=123456)
actual = [rrs.getnextblock() for _ in range(10)]
expected = test_data['raw-seed123456-index0-n10']
assert actual == expected
# Continuing the sequence, test conversion to floats
actual = [rrs.random() for _ in range(10)]
expected = test_data['double-seed123456-index10-n10']
assert actual == expected
# Continuing the sequence, test byte sequences
actual = [rrs.randbytes(10) for _ in range(2)]
expected = [
bytes(x) for x in test_data['bytes-len10-seed123456-index20-n2']
]
assert actual == expected
# Continuing the sequence, test double sequences
with rrs.cascade():
seq1 = [rrs.random() for _ in range(5)]
with rrs.cascade():
seq2 = [rrs.random() for _ in range(5)]
actual = [seq1, seq2]
expected = test_data['doubles-len5-seed123456-index22-n2']
assert actual == expected
# Continuing the sequence, test nil-kappa von mises dist
actual = [rrs.vonmisesvariate(mu=0.0, kappa=0.0) for _ in range(5)]
expected = test_data['vonmises-kappa0-mu0-seed123456-index24-n5']
assert actual == expected
# Continuing the sequence, test wrapped-mu von mises dist
actual = [rrs.vonmisesvariate(mu=10.0, kappa=2.0) for _ in range(5)]
expected = test_data['vonmises-kappa2-mu10-seed123456-index29-n5']
assert actual == expected
# Continuing the sequence, test large-alpha gamma dist
actual = [rrs.gammavariate(alpha=2.0, beta=1.0) for _ in range(5)]
expected = test_data['gamma-alpha2-beta1-seed123456-index34-n5']
assert actual == expected
# Continuing the sequence, test unit-alpha gamma dist
actual = [rrs.gammavariate(alpha=1.0, beta=1.0) for _ in range(5)]
expected = test_data['gamma-alpha1-beta1-seed123456-index39-n5']
assert actual == expected
# Continuing the sequence, test small-alpha gamma dist
actual = [rrs.gammavariate(alpha=0.5, beta=1.0) for _ in range(5)]
expected = test_data['gamma-alpha0.5-beta1-seed123456-index44-n5']
assert actual == expected
# Continuing the sequence, gaussian distribution
actual = [rrs.gauss(mu=0, sigma=1) for _ in range(10)]
expected = test_data['gauss-mu0-sigma1-seed123456-index49-n10']
assert actual == expected
def test_pareto_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.paretovariate(0.0)
def test_gamma_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.gammavariate(-1.0, -1.0)
def test_uniformproduct_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.uniformproduct(0)
def test_finitegeometric_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(ValueError):
rrs.finitegeometric(0.5, 0)
def test_choice_args():
rrs = samplespace.RepeatableRandomSequence()
with pytest.raises(IndexError):
rrs.choice([])
