def test_cant_get_autocorrelation_of_np_ints() -> None:
x = [1, 2, 3]
x_list = [np.array(a, int) for a in x]
with pytest.raises(
ValueError,
match=
"Autocorrelation must be run on a list of numpy floating types."):
stats.autocorrelation(x_list)
def test_can_get_correct_autocorrelation_when_nonscalar_input() -> None:
x_list = [
np.array([[1., 27.2], [10., 3.2]]),
np.array([[2., 99.4], [5., 4.6]]),
np.array([[3., 31.5], [10., 7.8]]),
np.array([[4., 14.3], [5., 2.1]]),
]
expected = np.array([[[1., 0.25, -0.3, -0.45], [1., -0.27679699, -0.32759603, 0.10439302]],
[[1., -0.75, 0.5, -0.25], [1., -0.40761833, -0.24778339, 0.15540172]]])
actual = np.array([[stats.autocorrelation(x_list, (0, 0)),
stats.autocorrelation(x_list, (0, 1))],
[stats.autocorrelation(x_list, (1, 0)),
stats.autocorrelation(x_list, (1, 1))]])
np.testing.assert_almost_equal(actual, expected)
def plot_acf(data: List[primitive_types], nlags: int = None) -> Any:
autocorr = stats.autocorrelation(data)
fig, ax = __create_new_mpl()
if nlags is None:
nlags = __calc_max_lag(len(autocorr))
__plot_corr(ax, autocorr, nlags)
return fig
def test_can_get_correct_autocorrelation() -> None:
x: List[primitive_types] = [1., 2., 3., 4., 5., 6., 7., 8.]
autocorr = stats.autocorrelation(x)
expected = [
1., 0.625, 0.27380952, -0.0297619, -0.26190476, -0.39880952,
-0.41666667, -0.29166667
]
np.testing.assert_almost_equal(autocorr, expected)
def test_can_get_correct_autocorrelation() -> None:
x = [1, 2, 3, 4, 5, 6, 7, 8]
x_list = [np.array(a, float) for a in x]
autocorr = stats.autocorrelation(x_list)
expected = [
1., 0.625, 0.27380952, -0.0297619, -0.26190476, -0.39880952,
-0.41666667, -0.29166667
]
np.testing.assert_almost_equal(autocorr, expected)
def plot_acf(data: List[numpy_types],
index: Tuple[int, ...] = (),
nlags: int = None) -> Any:
autocorr = stats.autocorrelation(data, index)
fig, ax = __create_new_mpl()
if nlags is None:
nlags = __calc_max_lag(len(autocorr))
__plot_corr(ax, autocorr, nlags)
return fig
def test_autocorrelation_same_for_streaming_as_batch() -> None:
with Model() as model:
model.uniform = Uniform(0, 1000)
net = model.to_bayes_net()
draws = 15
set_starting_state(model)
samples = sample(net=net, sample_from=net.get_latent_vertices(), algo="metropolis", draws=draws)
set_starting_state(model)
iter_samples = generate_samples(net=net, sample_from=net.get_latent_vertices(), algo="metropolis")
samples_dataframe = pd.DataFrame()
for next_sample in islice(iter_samples, draws):
samples_dataframe = samples_dataframe.append(next_sample, ignore_index=True)
for vertex_id in samples_dataframe:
autocorr_streaming = stats.autocorrelation(list(samples_dataframe[vertex_id].values))
autocorr_batch = stats.autocorrelation(samples[vertex_id])
np.testing.assert_array_equal(autocorr_batch, autocorr_streaming)
def test_autocorr_returns_ndarray_of_correct_dtype() -> None:
with Model() as m:
m.uniform = Uniform(0, 1000)
net = m.to_bayes_net()
samples = sample(net=net, sample_from=net.iter_latent_vertices(), draws=10)
valid_key = list(samples.keys())[0]
sample_ = samples.get(valid_key)
assert sample_ is not None
autocorr = stats.autocorrelation(sample_)
assert type(autocorr) == np.ndarray
def autocorrelation_example_nd():
with Model() as m:
m.a = Gaussian(np.array([[20., 30.], [40., 60.]]),
np.array([[1., 1.], [1., 1.]]))
bayes_net = m.to_bayes_net()
# %%SNIPPET_START%% PythonNdAutocorrelation
algo = MetropolisHastingsSampler()
posterior_samples = sample(net=bayes_net,
sample_from=bayes_net.get_latent_vertices(),
sampling_algorithm=algo,
draws=100)
vertex_samples = posterior_samples.get('a')
ac = stats.autocorrelation(vertex_samples, (0, 1))
def autocorrelation_example_scalar():
with Model() as m:
m.a = Gaussian(20, 1.)
m.b = Gaussian(20, 1.)
m.c = Gaussian(m.a + m.b, 1.)
m.c.observe(43.)
m.a.set_value(20.)
m.b.set_value(20.)
bayes_net = m.to_bayes_net()
# %%SNIPPET_START%% PythonScalarAutocorrelation
algo = MetropolisHastingsSampler()
posterior_samples = sample(net=bayes_net,
sample_from=bayes_net.get_latent_vertices(),
sampling_algorithm=algo,
draws=100)
vertex_samples = posterior_samples.get('a')
ac = stats.autocorrelation(vertex_samples)
def test_cant_get_autocorrelation_of_np_ints() -> None:
x: List[primitive_types] = [1, 2, 3]
with pytest.raises(
ValueError,
match=r"Autocorrelation must be run on a list of floating types"):
stats.autocorrelation(x)