def test_concatenate(self):
values = np.random.randn(20).reshape(4, 5)
stats_list = [stats.DataStats(float(x.mean()), float(x.var()), x.min(), x.max(), x.size) for x in values]
concatenated = stats.DataStats.concatenate(stats_list)
assert concatenated.mean == pytest.approx(np.mean(values))
assert concatenated.var == pytest.approx(np.var(values))
assert concatenated.min == pytest.approx(np.min(values))
assert concatenated.max == pytest.approx(np.max(values))
assert concatenated.num == values.size
def stats_per_utterance(self):
"""
Return statistics calculated for all samples of each utterance in the corpus.
Returns:
dict: A dictionary containing a DataStats object for each utt.
"""
all_stats = {}
for utterance in self.utterances.values():
data = utterance.read_samples()
all_stats[utterance.idx] = stats.DataStats(float(np.mean(data)),
float(np.var(data)),
np.min(data),
np.max(data), data.size)
return all_stats
def stats_per_key(self):
"""
Return statistics calculated for each key in the container.
Note:
The feature container has to be opened in advance.
Returns:
dict: A dictionary containing a DataStats object for each key.
"""
self.raise_error_if_not_open()
all_stats = {}
for key, data in self._file.items():
data = data[()]
all_stats[key] = stats.DataStats(float(np.mean(data)),
float(np.var(data)), np.min(data),
np.max(data), data.size)
return all_stats
def stats_per_utterance(self):
"""
Return statistics calculated for each utterance in the container.
Note:
The feature container has to be opened in advance.
Returns:
dict: A dictionary containing a DataStats object for each utterance.
"""
self._check_is_open()
all_stats = {}
for utt_id, data in self._file.items():
data = data[()]
all_stats[utt_id] = stats.DataStats(float(np.mean(data)),
float(np.var(data)),
np.min(data), np.max(data),
data.size)
return all_stats
def test_values(self):
s = stats.DataStats(2.3, 1.2, 4.0, -2, 99)
assert np.array_equal(s.values, np.array([2.3, 1.2, 4.0, -2, 99]))
def test_to_dict(self):
s = stats.DataStats(2.3, 1.2, -2, 4.0, 99)
d = s.to_dict()
assert d == {'mean': 2.3, 'var': 1.2, 'max': 4.0, 'min': -2, 'num': 99}