def test_mmotifs_two_motif_pairs_max_motifs_2(T):
motif_distances_ref = np.array(
[[0.00000000e00, 1.11510080e-07], [1.68587394e-07, 2.58694429e-01]]
)
motif_indices_ref = np.array([[2, 9], [6, 1]])
motif_subspaces_ref = [np.array([1]), np.array([2])]
motif_mdls_ref = [
np.array([232.0, 250.57542476, 260.0, 271.3509059]),
np.array([264.0, 280.0, 299.01955001, 310.51024953]),
]
m = 4
excl_zone = int(np.ceil(m / config.STUMPY_EXCL_ZONE_DENOM))
P, I = naive.mstump(T, m, excl_zone)
(
motif_distances_cmp,
motif_indices_cmp,
motif_subspaces_cmp,
motif_mdls_cmp,
) = mmotifs(
T, P, I, cutoffs=np.inf, max_motifs=2, max_distance=np.inf, max_matches=2
)
npt.assert_array_almost_equal(motif_distances_ref, motif_distances_cmp)
npt.assert_array_almost_equal(motif_indices_ref, motif_indices_cmp)
npt.assert_array_almost_equal(motif_subspaces_ref, motif_subspaces_cmp)
npt.assert_array_almost_equal(motif_mdls_ref, motif_mdls_cmp)
def test_mstump_discords(T, m):
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone, discords=True)
right_P, right_I = mstump(T, m, discords=True)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_mstump(T, m):
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I, _ = naive.mstump(T, m, excl_zone)
comp_P, comp_I = mstump(T, m)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mstump_discords(T, m):
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I = naive.mstump(T, m, excl_zone, discords=True)
comp_P, comp_I = mstump(T, m, discords=True)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mstumped_discords(T, m, dask_cluster):
with Client(dask_cluster) as dask_client:
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I = naive.mstump(T, m, excl_zone, discords=True)
comp_P, comp_I = mstumped(dask_client, T, m, discords=True)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mstumped_discords(T, m, dask_cluster):
with Client(dask_cluster) as dask_client:
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone, discords=True)
right_P, right_I = mstumped(dask_client, T, m, discords=True)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_mstumped_df(T, m, dask_cluster):
with Client(dask_cluster) as dask_client:
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I, _ = naive.mstump(T, m, excl_zone)
df = pd.DataFrame(T.T)
comp_P, comp_I = mstumped(dask_client, df, m)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mstumped_df(T, m, dask_cluster):
with Client(dask_cluster) as dask_client:
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone)
df = pd.DataFrame(T.T)
right_P, right_I = mstumped(dask_client, df, m)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_mstump_include(T, m):
for width in range(T.shape[0]):
for i in range(T.shape[0] - width):
include = np.asarray(range(i, i + width + 1))
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone, include)
right_P, right_I = mstump(T, m, include)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_constant_subsequence_self_join():
T_A = np.concatenate((np.zeros(20, dtype=np.float64), np.ones(5, dtype=np.float64)))
T = np.array([T_A, T_A, np.random.rand(T_A.shape[0])])
m = 3
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I, _ = naive.mstump(T, m, excl_zone)
comp_P, comp_I = mstump(T, m)
npt.assert_almost_equal(ref_P, comp_P) # ignore indices
def test_mstump_include_discords(T, m):
for width in range(T.shape[0]):
for i in range(T.shape[0] - width):
include = np.asarray(range(i, i + width + 1))
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I, _ = naive.mstump(T, m, excl_zone, include, discords=True)
comp_P, comp_I = mstump(T, m, include, discords=True)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_query_mstump_profile(T, m):
excl_zone = int(np.ceil(m / 4))
for query_idx in range(T.shape[0] - m + 1):
ref_P, ref_I = naive.mstump(T, m, excl_zone)
ref_P = ref_P[:, query_idx]
ref_I = ref_I[:, query_idx]
M_T, Σ_T = core.compute_mean_std(T, m)
comp_P, comp_I = _query_mstump_profile(query_idx, T, T, m, excl_zone,
M_T, Σ_T, M_T, Σ_T)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_equal(ref_I, comp_I)
def test_mstumped_one_subsequence_nan_self_join_all_dimensions(
T, m, substitution_location, dask_cluster):
with Client(dask_cluster) as dask_client:
excl_zone = int(np.ceil(m / 4))
T_sub = T.copy()
T_sub[:, substitution_location] = np.nan
ref_P, ref_I, _ = naive.mstump(T_sub, m, excl_zone)
comp_P, comp_I = mstumped(dask_client, T_sub, m)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mstumped_one_subsequence_inf_self_join_all_dimensions(
T, m, substitution_location, dask_cluster):
with Client(dask_cluster) as dask_client:
excl_zone = int(np.ceil(m / 4))
T_sub = T.copy()
T_sub[:, substitution_location] = np.inf
left_P, left_I = naive.mstump(T_sub, m, excl_zone)
right_P, right_I = mstumped(dask_client, T_sub, m)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_constant_subsequence_self_join(dask_cluster):
with Client(dask_cluster) as dask_client:
T_A = np.concatenate(
(np.zeros(20, dtype=np.float64), np.ones(5, dtype=np.float64)))
T = np.array([T_A, T_A, np.random.rand(T_A.shape[0])])
m = 3
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone)
right_P, right_I = mstumped(dask_client, T, m)
npt.assert_almost_equal(left_P, right_P) # ignore indices
def test_mstumped_include(T, m, dask_cluster):
with Client(dask_cluster) as dask_client:
for width in range(T.shape[0]):
for i in range(T.shape[0] - width):
include = np.asarray(range(i, i + width + 1))
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone, include)
right_P, right_I = mstumped(dask_client, T, m, include)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_naive_mstump():
T = np.random.uniform(-1000, 1000, [1, 1000]).astype(np.float64)
m = 20
zone = int(np.ceil(m / 4))
left = naive.stamp(T[0], m, exclusion_zone=zone)
left_P = left[np.newaxis, :, 0].T
left_I = left[np.newaxis, :, 1].T
right_P, right_I = naive.mstump(T, m, zone)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_get_first_mstump_profile(T, m):
excl_zone = int(np.ceil(m / 4))
start = 0
ref_P, ref_I = naive.mstump(T, m, excl_zone)
ref_P = ref_P[start, :]
ref_I = ref_I[start, :]
M_T, Σ_T = core.compute_mean_std(T, m)
comp_P, comp_I = _get_first_mstump_profile(start, T, T, m, excl_zone, M_T,
Σ_T, M_T, Σ_T)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_equal(ref_I, comp_I)
def test_mstump_nan_self_join_all_dimensions(T, m, substitute, substitution_locations):
excl_zone = int(np.ceil(m / 4))
T_sub = T.copy()
for substitution_location in substitution_locations:
T_sub[:] = T[:]
T_sub[:, substitution_location] = substitute
ref_P, ref_I, _ = naive.mstump(T_sub, m, excl_zone)
comp_P, comp_I = mstump(T_sub, m)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_naive_mstump():
T = np.random.uniform(-1000, 1000, [1, 1000]).astype(np.float64)
m = 20
zone = int(np.ceil(m / 4))
ref_mp = naive.stamp(T[0], m, exclusion_zone=zone)
ref_P = ref_mp[np.newaxis, :, 0].T
ref_I = ref_mp[np.newaxis, :, 1].T
comp_P, comp_I = naive.mstump(T, m, zone)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mstump_wrapper(T, m):
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I = naive.mstump(T, m, excl_zone)
comp_P, comp_I = mstump(T, m)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
df = pd.DataFrame(T.T)
comp_P, comp_I = mstump(df, m)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_get_first_mstump_profile(T, m):
excl_zone = int(np.ceil(m / 4))
start = 0
left_P, left_I = naive.mstump(T, m, excl_zone)
left_P = left_P[start, :]
left_I = left_I[start, :]
M_T, Σ_T = core.compute_mean_std(T, m)
right_P, right_I = _get_first_mstump_profile(start, T, T, m, excl_zone,
M_T, Σ_T, M_T, Σ_T)
npt.assert_almost_equal(left_P, right_P)
npt.assert_equal(left_I, right_I)
def test_mstump_wrapper(T, m):
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone)
right_P, right_I = mstump(T, m)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
df = pd.DataFrame(T.T)
right_P, right_I = mstump(df, m)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_mstump_nan_inf_self_join_first_dimension(T, m, substitute,
substitution_locations):
excl_zone = int(np.ceil(m / 4))
T_sub = T.copy()
for substitution_location in substitution_locations:
T_sub[:] = T[:]
T_sub[0, substitution_location] = substitute
left_P, left_I = naive.mstump(T_sub, m, excl_zone)
right_P, right_I = mstump(T_sub, m)
npt.assert_almost_equal(left_P, right_P)
npt.assert_almost_equal(left_I, right_I)
def test_identical_subsequence_self_join():
identical = np.random.rand(8)
T_A = np.random.rand(20)
T_A[1:1 + identical.shape[0]] = identical
T_A[11:11 + identical.shape[0]] = identical
T = np.array([T_A, T_A, np.random.rand(T_A.shape[0])])
m = 3
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone)
right_P, right_I = mstump(T, m)
npt.assert_almost_equal(left_P, right_P,
decimal=naive.PRECISION) # ignore indices
def test_identical_subsequence_self_join():
identical = np.random.rand(8)
T_A = np.random.rand(20)
T_A[1:1 + identical.shape[0]] = identical
T_A[11:11 + identical.shape[0]] = identical
T = np.array([T_A, T_A, np.random.rand(T_A.shape[0])])
m = 3
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I = naive.mstump(T, m, excl_zone)
comp_P, comp_I = mstump(T, m)
npt.assert_almost_equal(
ref_P, comp_P, decimal=config.STUMPY_TEST_PRECISION) # ignore indices
def test_mstumped_identical_subsequence_self_join(dask_cluster):
with Client(dask_cluster) as dask_client:
identical = np.random.rand(8)
T_A = np.random.rand(20)
T_A[1:1 + identical.shape[0]] = identical
T_A[11:11 + identical.shape[0]] = identical
T = np.array([T_A, T_A, np.random.rand(T_A.shape[0])])
m = 3
excl_zone = int(np.ceil(m / 4))
left_P, left_I = naive.mstump(T, m, excl_zone)
right_P, right_I = mstumped(dask_client, T, m)
npt.assert_almost_equal(
left_P, right_P,
decimal=config.STUMPY_TEST_PRECISION) # ignore indices
def test_mstump_wrapper_include(T, m):
for width in range(T.shape[0]):
for i in range(T.shape[0] - width):
include = np.asarray(range(i, i + width + 1))
excl_zone = int(np.ceil(m / 4))
ref_P, ref_I = naive.mstump(T, m, excl_zone, include)
comp_P, comp_I = mstump(T, m, include)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
df = pd.DataFrame(T.T)
comp_P, comp_I = mstump(df, m, include)
npt.assert_almost_equal(ref_P, comp_P)
npt.assert_almost_equal(ref_I, comp_I)
def test_mmotifs_max_matches_none(T):
motif_distances_ref = np.array([[0.0000000e00, 1.1151008e-07]])
motif_indices_ref = np.array([[2, 9]])
motif_subspaces_ref = [np.array([1])]
motif_mdls_ref = [np.array([232.0, 250.57542476, 260.0, 271.3509059])]
m = 4
excl_zone = int(np.ceil(m / config.STUMPY_EXCL_ZONE_DENOM))
P, I = naive.mstump(T, m, excl_zone)
(
motif_distances_cmp,
motif_indices_cmp,
motif_subspaces_cmp,
motif_mdls_cmp,
) = mmotifs(T, P, I, max_matches=None)
npt.assert_array_almost_equal(motif_distances_ref, motif_distances_cmp)
npt.assert_array_almost_equal(motif_indices_ref, motif_indices_cmp)
npt.assert_array_almost_equal(motif_subspaces_ref, motif_subspaces_cmp)
npt.assert_array_almost_equal(motif_mdls_ref, motif_mdls_cmp)
def test_mmotifs_max_matches_2_k_1(T):
motif_distances_ref = np.array([[0.0, 0.20948156]])
motif_indices_ref = np.array([[2, 9]])
motif_subspaces_ref = [np.array([1, 3])]
motif_mdls_ref = [np.array([232.0, 250.57542476, 260.0, 271.3509059])]
m = 4
excl_zone = int(np.ceil(m / config.STUMPY_EXCL_ZONE_DENOM))
P, I = naive.mstump(T, m, excl_zone)
(
motif_distances_cmp,
motif_indices_cmp,
motif_subspaces_cmp,
motif_mdls_cmp,
) = mmotifs(T, P, I, max_distance=np.inf, max_matches=2, k=1)
npt.assert_array_almost_equal(motif_distances_ref, motif_distances_cmp)
npt.assert_array_almost_equal(motif_indices_ref, motif_indices_cmp)
npt.assert_array_almost_equal(motif_subspaces_ref, motif_subspaces_cmp)
npt.assert_array_almost_equal(motif_mdls_ref, motif_mdls_cmp)
