def test_enforce_noshuffle(self):
"""Test when we enforce uniqueness, but do not shuffle."""
(concatenated_train_sequence,
concatenated_train_cluster_id) = utils.concatenate_training_data(
self.train_sequences, self.train_cluster_ids, True, False)
self.assertListEqual([0.0] * 6 + [1.0] * 8,
concatenated_train_sequence.flatten().tolist())
self.assertEqual(7, len(concatenated_train_cluster_id))
self.assertEqual(5, len(set(concatenated_train_cluster_id)))
def test_noenforce_noshuffle(self):
"""Test when we do not enforce uniqueness, and do not shuffle."""
(concatenated_train_sequence,
concatenated_train_cluster_id) = utils.concatenate_training_data(
self.train_sequences, self.train_cluster_ids, False, False)
self.assertListEqual([0.0] * 6 + [1.0] * 8,
concatenated_train_sequence.flatten().tolist())
self.assertListEqual(['a', 'b', 'a', 'a', 'b', 'c', 'b'],
concatenated_train_cluster_id)
def fit(self, train_sequences, train_cluster_ids, args):
"""Fit UISRNN model.
Args:
train_sequences: Either a list of training sequences, or a single
concatenated training sequence:
(1) train_sequences is list, and each element is a 2-dim numpy array
of real numbers, of size size: length * D.
The length varies among differnt sequences, but the D is the same.
In speaker diarization, each sequence is the sequence of speaker
embeddings of one utterance.
(2) train_sequences is a single concatenated sequence, which is a
2-dim numpy array of real numbers. See _fit_concatenated()
for more details.
train_cluster_ids: Ground truth labels for train_sequences:
(1) if train_sequences is a list, this must also be a list of the same
size, each element being a 1-dim list or numpy array of strings.
(2) if train_sequences is a single concatenated sequence, this
must also be the concatenated 1-dim list or numpy array of strings
args: Training configurations. See arguments.py for details.
Raises:
TypeError: If train_sequences or train_cluster_ids is of wrong type.
"""
if isinstance(train_sequences, np.ndarray):
# train_sequences is already the concatenated sequence
concatenated_train_sequence = train_sequences
concatenated_train_cluster_id = train_cluster_ids
elif isinstance(train_sequences, list):
# train_sequences is a list of un-concatenated sequences,
# then we concatenate them first
(concatenated_train_sequence,
concatenated_train_cluster_id) = utils.concatenate_training_data(
train_sequences, train_cluster_ids,
args.enforce_cluster_id_uniqueness, True)
else:
raise TypeError('train_sequences must be a list or numpy.ndarray')
self._fit_concatenated(concatenated_train_sequence,
concatenated_train_cluster_id, args)
def test_noenforce_shuffle(self):
"""Test when we do not enforce uniqueness, but do shuffle."""
(concatenated_train_sequence,
concatenated_train_cluster_id) = utils.concatenate_training_data(
self.train_sequences, self.train_cluster_ids, False, True)
try:
self.assertListEqual(
[0.0] * 6 + [1.0] * 8,
concatenated_train_sequence.flatten().tolist())
self.assertListEqual(['a', 'b', 'a', 'a', 'b', 'c', 'b'],
concatenated_train_cluster_id)
except AssertionError:
self.assertListEqual(
[1.0] * 8 + [0.0] * 6,
concatenated_train_sequence.flatten().tolist())
self.assertListEqual(['a', 'b', 'c', 'b', 'a', 'b', 'a'],
concatenated_train_cluster_id)
def fit(self, train_sequences, train_cluster_ids, args):
"""Fit UISRNN model.
Args:
train_sequences: Either a list of training sequences, or a single
concatenated training sequence:
1. train_sequences is list, and each element is a 2-dim numpy array
of real numbers, of size: `length * D`.
The length varies among different sequences, but the D is the same.
In speaker diarization, each sequence is the sequence of speaker
embeddings of one utterance.
2. train_sequences is a single concatenated sequence, which is a
2-dim numpy array of real numbers. See `fit_concatenated()`
for more details.
train_cluster_ids: Ground truth labels for train_sequences:
1. if train_sequences is a list, this must also be a list of the same
size, each element being a 1-dim list or numpy array of strings.
2. if train_sequences is a single concatenated sequence, this
must also be the concatenated 1-dim list or numpy array of strings
args: Training configurations. See `arguments.py` for details.
Raises:
TypeError: If train_sequences or train_cluster_ids is of wrong type.
"""
if isinstance(train_sequences, np.ndarray):
# train_sequences is already the concatenated sequence
if self.estimate_transition_bias:
# see issue #55: https://github.com/google/uis-rnn/issues/55
self.logger.print(
2,
'Warning: transition_bias cannot be correctly estimated from a '
'concatenated sequence; train_sequences will be treated as a '
'single sequence. This can lead to inaccurate estimation of '
'transition_bias. Please, consider estimating transition_bias '
'before concatenating the sequences and passing it as argument.'
)
train_sequences = [train_sequences]
train_cluster_ids = [train_cluster_ids]
elif isinstance(train_sequences, list):
# train_sequences is a list of un-concatenated sequences
# we will concatenate it later, after estimating transition_bias
pass
else:
raise TypeError('train_sequences must be a list or numpy.ndarray')
# estimate transition_bias
if self.estimate_transition_bias:
(transition_bias, transition_bias_denominator
) = utils.estimate_transition_bias(train_cluster_ids)
# set or update transition_bias
if self.transition_bias is None:
self.transition_bias = transition_bias
self.transition_bias_denominator = transition_bias_denominator
else:
self.transition_bias = (
self.transition_bias * self.transition_bias_denominator +
transition_bias * transition_bias_denominator) / (
self.transition_bias_denominator +
transition_bias_denominator)
self.transition_bias_denominator += transition_bias_denominator
# concatenate train_sequences
(concatenated_train_sequence,
concatenated_train_cluster_id) = utils.concatenate_training_data(
train_sequences, train_cluster_ids,
args.enforce_cluster_id_uniqueness, True)
self.fit_concatenated(concatenated_train_sequence,
concatenated_train_cluster_id, args)
if i <= train_cases:
train_cluster_id.append(train_clus[j])
else:
train_cluster_id.append(list(map(int, train_clus[j])))
if False:
print('Processed case:', case_id)
print('emb shape:', np.shape(train_seq))
print('label shape:', np.shape(train_clus))
print('emb len:', len(train_sequence))
print('label len:', len(train_cluster_id))
if i <= train_cases:
trn_seq_lst.append(train_sequence)
trn_cluster_lst.append(train_cluster_id)
else:
test_seq_lst.append(train_sequence)
test_cluster_lst.append(train_cluster_id)
from uisrnn import utils
item = 0
(concatenated_train_sequence,
concatenated_train_cluster_id) = utils.concatenate_training_data(
trn_seq_lst[item], trn_cluster_lst[item], False, False)
if verbose:
print(type(concatenated_train_sequence),
type(concatenated_train_sequence[0]))
#print(np.shape(concatenated_train_sequence))
print(np.shape(concatenated_train_sequence))
