def test_large_call():
""" Test calling twed """
# Call TWED
dist = twed(A, TA, B, TB, nu, lamb, degree)
print('Python cuTWED distance: {:f}'.format(dist))
assert np.allclose(dist, reference_result)
def test_basic_call_float():
""" Test the same call in single precision by feeding different types. """
dist = twed(A.astype(np.float32), TA.astype(np.float32),
B.astype(np.float32), TB.astype(np.float32), nu, lamb, degree)
print('Python cuTWED distance (single precision): {:f}'.format(dist))
assert np.allclose(dist, reference_result)
def test_multi_twed():
D = np.zeros(sz)
print("Computing ctwed batch")
for row, A in enumerate(AA):
for col, B in enumerate(BB):
if col < row:
continue
dist = twed(A, T, B, T, nu, lamb, degree)
D[row][col] = dist
assert np.allclose(D, DIST)
def test_cutwed():
""" Test running the synthetic control dataset"""
DistanceMatrix = np.zeros((nseries, nseries))
for row, A in enumerate(tqdm(TS)):
for col, B in enumerate(TS):
if col < row:
continue
dist = twed(A, T, B, T, nu, lamb, degree)
DistanceMatrix[row][col] = dist
# print(f'Python CTWED distance:\t{row}\t{col}\t{dist:f}')
name = 'synthetic_distance_matrix_cutwed'
with open(f'{name}.npy', 'wb') as fh:
np.save(fh, DistanceMatrix)
# with sns.axes_style("white"):
# sns.heatmap(DistanceMatrix, square=True, cmap="YlGnBu")
# plt.savefig(f'{name}.png')
return DistanceMatrix
def test_single_cutwed():
print("Computing cutwed")
dist = twed(A, T, B, T, nu, lamb, degree)
print(f"Distance: {dist}")
assert np.allclose(single_ref, dist)
plt.plot()
# In[9]:
DistanceMatrix_ref[0, -1]
# In[10]:
# Test running the synthetic index dataset"""
DistanceMatrix_cu = np.zeros((nseries, nseries))
for (row, A), (col, B) in tqdm(product(enumerate(TS), enumerate(TS)),
total=len(TS)**2):
if col < row:
continue
dist = twed(A, T, B, T, nu, lamb, degree)
DistanceMatrix_cu[row][col] = dist
with sns.axes_style("white"):
sns.heatmap(DistanceMatrix_cu, square=True, cmap="YlGnBu")
plt.plot()
# In[11]:
DistanceMatrix_cu[0, -1]
# In[12]:
np.max(np.abs(DistanceMatrix_cu - DistanceMatrix_ref))
# In[13]: