def _get_small_datasets(padded=False, duration=False, padded_length=1000):
if duration:
X, Y = example_file_data_sources_for_duration_model()
else:
X, Y = example_file_data_sources_for_acoustic_model()
if padded:
X = PaddedFileSourceDataset(X, padded_length=padded_length)
Y = PaddedFileSourceDataset(Y, padded_length=padded_length)
else:
X = FileSourceDataset(X)
Y = FileSourceDataset(Y)
return X, Y
def test_dtw_frame_length_adjastment():
_, X = example_file_data_sources_for_duration_model()
X = FileSourceDataset(X)
X_unaligned = X.asarray()
# This should trigger frame length adjastment
Y_unaligned = np.pad(X_unaligned, [(0, 0), (5, 0), (0, 0)],
mode="constant", constant_values=0)
Y_unaligned = Y_unaligned[:, :-5, :]
for aligner in [DTWAligner(), IterativeDTWAligner(
n_iter=1, max_iter_gmm=1, n_components_gmm=1)]:
X_aligned, Y_aligned = aligner.transform((X_unaligned, Y_unaligned))
assert X_aligned.shape == Y_aligned.shape