def test_add_sample(self):
mean_data = 1.0
count_data = 0
sum_of_sqd_deviations = 2
counts_per_bucket = [1, 1, 1, 1]
bounds = [0.5, 1, 1.5]
value = 3
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.add_sample(value, None, None)
self.assertEqual(1, dist_agg_data.count_data)
self.assertEqual(value, dist_agg_data.mean_data)
count_data = 1
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.add_sample(value, None, None)
self.assertEqual(2, dist_agg_data.count_data)
self.assertEqual(2.0, dist_agg_data.mean_data)
self.assertEqual(4.0, dist_agg_data.sum_of_sqd_deviations)
self.assertIsNot(0, dist_agg_data.count_data)
def test_init_bad_bounds(self):
# Check that bounds are unique
with self.assertRaises(AssertionError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
min_=mock.Mock(),
max_=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=[0, 0, 0, 0],
bounds=[1, 2, 2])
# Check that bounds are sorted
with self.assertRaises(AssertionError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
min_=mock.Mock(),
max_=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=[0, 0, 0, 0],
bounds=[1, 3, 2])
# Check that all bounds are positive
with self.assertRaises(AssertionError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
min_=mock.Mock(),
max_=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=[0, 0, 0, 0],
bounds=[-1, 1, 2])
def test_variance(self):
mean_data = mock.Mock()
count_data = 0
_min = mock.Mock()
_max = mock.Mock()
sum_of_sqd_deviations = mock.Mock()
counts_per_bucket = [1, 1, 1]
bounds = [1.0 / 2.0, 1]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
self.assertEqual(0, dist_agg_data.variance)
count_data = 2
sum_of_sqd_deviations = 2
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
self.assertEqual(2.0, dist_agg_data.variance)
def test_init_bad_bucket_counts(self):
# Check that len(counts_per_bucket) == len(bounds) + 1
with self.assertRaises(AssertionError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
min_=mock.Mock(),
max_=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=[0, 0, 0],
bounds=[1, 2, 3])
# Check that counts aren't negative
with self.assertRaises(AssertionError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
min_=mock.Mock(),
max_=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=[0, 2, -2, 0],
bounds=[1, 2, 3])
# And check that we don't throw given the right args
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
min_=mock.Mock(),
max_=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=[0, 0, 0, 0],
bounds=[1, 2, 3])
def test_increment_bucket_count(self):
mean_data = mock.Mock()
count_data = mock.Mock()
_min = 0
_max = 1
sum_of_sqd_deviations = mock.Mock()
counts_per_bucket = [0]
bounds = []
value = 1
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_= _min,
max_ = _max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds
)
dist_agg_data.increment_bucket_count(value=value)
self.assertEqual([1], dist_agg_data.counts_per_bucket)
counts_per_bucket = [1, 1]
bounds = [1/4, 3/2]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_= _min,
max_ = _max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds
)
dist_agg_data.increment_bucket_count(value=value)
self.assertEqual([1, 2], dist_agg_data.counts_per_bucket)
bounds = [1/4, 1/2]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_= _min,
max_ = _max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds
)
dist_agg_data.increment_bucket_count(value=value)
self.assertEqual([1, 3], dist_agg_data.counts_per_bucket)
def test_add_sample_attachment(self):
mean_data = 1.0
count_data = 1
_min = 0
_max = 1
sum_of_sqd_deviations = 2
counts_per_bucket = [1, 1, 1, 1]
bounds = [0.5, 1, 1.5]
value = 3
timestamp = time.time()
attachments = {"One": "one", "Two": "two"}
exemplar_1 = aggregation_data_module.Exemplar(4, timestamp,
attachments)
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds,
exemplars=[None, None, None, exemplar_1])
self.assertEqual(dist_agg_data.exemplars[3], exemplar_1)
dist_agg_data.add_sample(value, timestamp, attachments)
self.assertEqual(0, dist_agg_data.min)
self.assertEqual(3, dist_agg_data.max)
self.assertEqual(2, dist_agg_data.count_data)
self.assertEqual(2.0, dist_agg_data.mean_data)
# Check that adding a sample overwrites the bucket's exemplar
self.assertNotEqual(dist_agg_data.exemplars[3], exemplar_1)
self.assertEqual(dist_agg_data.exemplars[3].value, 3)
self.assertEqual(dist_agg_data.exemplars[3].attachments, attachments)
count_data = 4
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=[2, 1, 2, 1, 1, 1],
bounds=[1, 2, 3, 4, 5])
dist_agg_data.add_sample(value, timestamp, attachments)
self.assertEqual(5, dist_agg_data.count_data)
self.assertEqual(1.4, dist_agg_data.mean_data)
self.assertEqual(5.2, dist_agg_data.sum_of_sqd_deviations)
self.assertIsNot(0, dist_agg_data.count_data)
self.assertEqual(3, dist_agg_data.exemplars[3].value)
def test_constructor(self):
mean_data = 1
count_data = 0
_min = 0
_max = 1
sum_of_sqd_deviations = mock.Mock()
counts_per_bucket = [1, 1, 1]
bounds = [1.0 / 2.0, 1]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
self.assertEqual(1, dist_agg_data.mean_data)
self.assertEqual(0, dist_agg_data.count_data)
self.assertEqual(0, dist_agg_data.min)
self.assertEqual(1, dist_agg_data.max)
self.assertEqual(sum_of_sqd_deviations,
dist_agg_data.sum_of_sqd_deviations)
self.assertEqual([1, 1, 1], dist_agg_data.counts_per_bucket)
self.assertEqual([1.0 / 2.0, 1], dist_agg_data.bounds)
self.assertIsNotNone(dist_agg_data.sum)
self.assertEqual(0, dist_agg_data.variance)
def test_to_point(self):
timestamp = datetime(1970, 1, 1)
ex_9 = aggregation_data_module.Exemplar(
9, timestamp, {'trace_id': 'dead', 'span_id': 'beef'}
)
ex_99 = aggregation_data_module.Exemplar(
99, timestamp, {'trace_id': 'dead', 'span_id': 'bef0'}
)
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=50,
count_data=99,
sum_of_sqd_deviations=80850.0,
counts_per_bucket=[0, 9, 90, 0],
bounds=[1, 10, 100],
exemplars=[None, ex_9, ex_99, None],
)
converted_point = dist_agg_data.to_point(timestamp)
self.assertTrue(isinstance(converted_point.value,
value.ValueDistribution))
self.assertEqual(converted_point.value.count, 99)
self.assertEqual(converted_point.value.sum, 4950)
self.assertEqual(converted_point.value.sum_of_squared_deviation,
80850.0)
self.assertEqual([bb.count for bb in converted_point.value.buckets],
[0, 9, 90, 0])
self.assertEqual(converted_point.value.bucket_options.type_.bounds,
[1, 10, 100])
self.assertTrue(
exemplars_equal(
ex_9,
converted_point.value.buckets[1].exemplar))
self.assertTrue(
exemplars_equal(
ex_99,
converted_point.value.buckets[2].exemplar))
def test_constructor_with_exemplar(self):
timestamp = time.time()
attachments = {"One": "one", "Two": "two"}
exemplars = [
aggregation_data_module.Exemplar(.07, timestamp, attachments),
aggregation_data_module.Exemplar(.7, timestamp, attachments),
aggregation_data_module.Exemplar(7, timestamp, attachments)
]
mean_data = 2.59
count_data = 3
sum_of_sqd_deviations = mock.Mock()
counts_per_bucket = [1, 1, 1]
bounds = [1.0 / 2.0, 1]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
sum_of_sqd_deviations=sum_of_sqd_deviations,
exemplars=exemplars,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
self.assertEqual(dist_agg_data.mean_data, mean_data)
self.assertEqual(dist_agg_data.count_data, count_data)
self.assertEqual(dist_agg_data.sum_of_sqd_deviations,
sum_of_sqd_deviations)
self.assertEqual(dist_agg_data.counts_per_bucket, counts_per_bucket)
self.assertEqual(dist_agg_data.bounds, bounds)
self.assertEqual(dist_agg_data.sum, mean_data * count_data)
for ii, ex in enumerate(exemplars):
self.assertEqual(dist_agg_data.exemplars[ii], ex)
def __init__(self,
boundaries=None,
distribution=None,
aggregation_type=Type.DISTRIBUTION):
if boundaries:
if not all(boundaries[ii] < boundaries[ii + 1]
for ii in range(len(boundaries) - 1)):
raise ValueError("bounds must be sorted in increasing order")
for ii, bb in enumerate(boundaries):
if bb > 0:
break
else:
ii += 1
if ii:
logger.warning("Dropping %s non-positive bucket boundaries",
ii)
boundaries = boundaries[ii:]
super(DistributionAggregation,
self).__init__(buckets=boundaries,
aggregation_type=aggregation_type)
self._boundaries = bucket_boundaries.BucketBoundaries(boundaries)
self._distribution = distribution or {}
self.aggregation_data = aggregation_data.DistributionAggregationData(
0, 0, 0, None, boundaries)
def test_add_sample(self):
mean_data = 1.0
count_data = 0
_min = 0
_max = 1
sum_of_sqd_deviations = 2
counts_per_bucket = [1, 1, 1, 1]
bounds = [0, 0.5, 1, 1.5]
value = 3
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.add_sample(value=value)
self.assertEqual(0, dist_agg_data.min)
self.assertEqual(3, dist_agg_data.max)
self.assertEqual(1, dist_agg_data.count_data)
self.assertEqual(value, dist_agg_data.mean_data)
count_data = 1
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.add_sample(value=value)
self.assertEqual(2, dist_agg_data.count_data)
self.assertEqual(2.0, dist_agg_data.mean_data)
self.assertEqual(4.0, dist_agg_data.sum_of_sqd_deviations)
self.assertIsNot(0, dist_agg_data.count_data)
value_2 = -1
dist_agg_data.add_sample(value=value_2)
self.assertEqual(value_2, dist_agg_data.min)
def test_increment_bucket_count(self):
mean_data = mock.Mock()
count_data = mock.Mock()
sum_of_sqd_deviations = mock.Mock()
counts_per_bucket = [0]
bounds = []
value = 1
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.increment_bucket_count(value=value)
self.assertEqual([1], dist_agg_data.counts_per_bucket)
counts_per_bucket = [1, 1, 1]
bounds = [1.0 / 4.0, 3.0 / 2.0]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.increment_bucket_count(value=value)
self.assertEqual([1, 2, 1], dist_agg_data.counts_per_bucket)
bounds = [1.0 / 4.0, 1.0 / 2.0]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
sum_of_sqd_deviations=sum_of_sqd_deviations,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
dist_agg_data.increment_bucket_count(value=value)
self.assertEqual([1, 2, 2], dist_agg_data.counts_per_bucket)
def __init__(self,
boundaries=None,
distribution=None,
aggregation_type=Type.DISTRIBUTION):
super(DistributionAggregation,
self).__init__(buckets=boundaries,
aggregation_type=aggregation_type)
self._boundaries = bucket_boundaries.BucketBoundaries(boundaries)
self._distribution = distribution or {}
self.aggregation_data = aggregation_data.DistributionAggregationData(
0, 0, float('inf'), float('-inf'), 0, None, boundaries)
def test_init_bad_exemplars(self):
# Check that we don't allow exemplars without bounds
with self.assertRaises(ValueError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=mock.Mock(),
bounds=None,
exemplars=[mock.Mock()])
# Check that the exemplar count matches the bucket count
with self.assertRaises(ValueError):
aggregation_data_module.DistributionAggregationData(
mean_data=mock.Mock(),
count_data=mock.Mock(),
sum_of_sqd_deviations=mock.Mock(),
counts_per_bucket=mock.Mock(),
bounds=[0, 1],
exemplars=[mock.Mock(), mock.Mock()])
def test_to_point_no_histogram(self):
timestamp = datetime(1970, 1, 1)
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=50,
count_data=99,
sum_of_sqd_deviations=80850.0,
)
converted_point = dist_agg_data.to_point(timestamp)
self.assertTrue(isinstance(converted_point.value,
value.ValueDistribution))
self.assertEqual(converted_point.value.count, 99)
self.assertEqual(converted_point.value.sum, 4950)
self.assertEqual(converted_point.value.sum_of_squared_deviation,
80850.0)
self.assertIsNone(converted_point.value.buckets)
self.assertIsNone(converted_point.value.bucket_options._type)
def test_create_timeseries_from_distribution(self):
"""Check for explicit 0-bound bucket for SD export."""
agg = aggregation_module.DistributionAggregation(
aggregation_type=aggregation_module.Type.DISTRIBUTION)
view = view_module.View(
name="example.org/test_view",
description="example.org/test_view",
columns=['tag_key'],
measure=mock.Mock(),
aggregation=agg,
)
v_data = view_data_module.ViewData(
view=view,
start_time=TEST_TIME_STR,
end_time=TEST_TIME_STR,
)
# Aggregation over (10 * range(10)) for buckets [2, 4, 6, 8]
dad = aggregation_data_module.DistributionAggregationData(
mean_data=4.5,
count_data=100,
sum_of_sqd_deviations=825,
counts_per_bucket=[20, 20, 20, 20, 20],
bounds=[2, 4, 6, 8],
exemplars={mock.Mock()
for ii in range(5)})
v_data._tag_value_aggregation_data_map = {('tag_value', ): dad}
v_data = metric_utils.view_data_to_metric(v_data, TEST_TIME)
exporter = stackdriver.StackdriverStatsExporter()
time_series_list = exporter.create_time_series_list(v_data)
self.assertEqual(len(time_series_list), 1)
[time_series] = time_series_list
self.check_labels(time_series.metric.labels, {'tag_key': 'tag_value'},
include_opencensus=True)
self.assertEqual(len(time_series.points), 1)
[point] = time_series.points
dv = point.value.distribution_value
self.assertEqual(100, dv.count)
self.assertEqual(825.0, dv.sum_of_squared_deviation)
self.assertEqual([0, 20, 20, 20, 20, 20], dv.bucket_counts)
self.assertEqual([0, 2, 4, 6, 8],
dv.bucket_options.explicit_buckets.bounds)
def test_create_timeseries_from_distribution(self):
"""Check for explicit 0-bound bucket for SD export."""
v_data = mock.Mock(spec=view_data_module.ViewData)
v_data.view.name = "example.org/test_view"
v_data.view.columns = ['tag_key']
v_data.view.aggregation.aggregation_type = \
aggregation_module.Type.DISTRIBUTION
v_data.start_time = TEST_TIME
v_data.end_time = TEST_TIME
# Aggregation over (10 * range(10)) for buckets [2, 4, 6, 8]
dad = aggregation_data_module.DistributionAggregationData(
mean_data=4.5,
count_data=100,
min_=0,
max_=9,
sum_of_sqd_deviations=825,
counts_per_bucket=[20, 20, 20, 20, 20],
bounds=[2, 4, 6, 8],
exemplars={mock.Mock()
for ii in range(5)})
v_data.tag_value_aggregation_data_map = {('tag_value', ): dad}
exporter = stackdriver.StackdriverStatsExporter(
options=mock.Mock(),
client=mock.Mock(),
)
time_series_list = exporter.create_time_series_list(v_data, "", "")
self.assertEqual(len(time_series_list), 1)
[time_series] = time_series_list
self.assertCorrectLabels(time_series.metric.labels,
{'tag_key': 'tag_value'},
include_opencensus=True)
self.assertEqual(len(time_series.points), 1)
[point] = time_series.points
dv = point.value.distribution_value
self.assertEqual(100, dv.count)
self.assertEqual(4.5, dv.mean)
self.assertEqual(825.0, dv.sum_of_squared_deviation)
self.assertEqual([0, 20, 20, 20, 20, 20], dv.bucket_counts)
self.assertEqual([0, 2, 4, 6, 8],
dv.bucket_options.explicit_buckets.bounds)
def test_constructor_with_exemplar(self):
timestamp = time.time()
attachments = {"One": "one", "Two": "two"}
exemplar_1 = aggregation_data_module.Exemplar(4, timestamp,
attachments)
exemplar_2 = aggregation_data_module.Exemplar(5, timestamp,
attachments)
mean_data = 1
count_data = 0
_min = 0
_max = 1
sum_of_sqd_deviations = mock.Mock()
counts_per_bucket = [1, 1, 1, 1]
bounds = [0, 1.0 / 2.0, 1]
exemplars = [exemplar_1, exemplar_2]
dist_agg_data = aggregation_data_module.DistributionAggregationData(
mean_data=mean_data,
count_data=count_data,
min_=_min,
max_=_max,
sum_of_sqd_deviations=sum_of_sqd_deviations,
exemplars=exemplars,
counts_per_bucket=counts_per_bucket,
bounds=bounds)
self.assertEqual(1, dist_agg_data.mean_data)
self.assertEqual(0, dist_agg_data.count_data)
self.assertEqual(0, dist_agg_data.min)
self.assertEqual(1, dist_agg_data.max)
self.assertEqual(sum_of_sqd_deviations,
dist_agg_data.sum_of_sqd_deviations)
self.assertEqual([1, 1, 1, 1], dist_agg_data.counts_per_bucket)
self.assertEqual([exemplar_1, exemplar_2], dist_agg_data.exemplars[3])
self.assertEqual([0, 1.0 / 2.0, 1], dist_agg_data.bounds)
self.assertIsNotNone(dist_agg_data.sum)
self.assertEqual(0, dist_agg_data.variance)
def test_create_timeseries_from_distribution(self):
"""Check for explicit 0-bound bucket for SD export."""
v_data = mock.Mock(spec=view_data_module.ViewData)
v_data.view.name = "example.org/test_view"
v_data.start_time = '2018-11-21T00:12:34.56Z'
v_data.end_time = '2018-11-21T00:23:45.67Z'
# Aggregation over (10 * range(10)) for buckets [2, 4, 6, 8]
dad = aggregation_data_module.DistributionAggregationData(
mean_data=4.5,
count_data=100,
min_=0,
max_=9,
sum_of_sqd_deviations=825,
counts_per_bucket=[20, 20, 20, 20, 20],
bounds=[2, 4, 6, 8],
exemplars={mock.Mock()
for ii in range(5)})
v_data.tag_value_aggregation_data_map = ({'tag_key': dad})
exporter = stackdriver.StackdriverStatsExporter(
options=mock.Mock(),
client=mock.Mock(),
)
time_series = exporter.create_time_series_list(v_data, "", "")
self.assertEqual(len(time_series.points), 1)
[point] = time_series.points
dv = point.value.distribution_value
self.assertEqual(100, dv.count)
self.assertEqual(4.5, dv.mean)
self.assertEqual(825.0, dv.sum_of_squared_deviation)
self.assertEqual([0, 20, 20, 20, 20, 20], dv.bucket_counts)
self.assertEqual([0, 2, 4, 6, 8],
dv.bucket_options.explicit_buckets.bounds)
def new_aggregation_data(self, measure=None):
"""Get a new AggregationData for this aggregation."""
return aggregation_data.DistributionAggregationData(
0, 0, 0, None, self._boundaries)
