def expect_adds_test_samples_ok(self, make_dist_func):
for t in _TEST_SAMPLES_AND_BUCKETS:
d = make_dist_func()
samples = t[u'samples']
for s in samples:
distribution.add_sample(s, d)
expect(d.bucketCounts).to(equal(t[u'want']))
_expect_stats_eq_direct_calc_from_samples(d, samples)
def expect_adds_test_samples_ok(self, make_dist_func):
for t in _TEST_SAMPLES_AND_BUCKETS:
d = make_dist_func()
samples = t[u'samples']
for s in samples:
distribution.add_sample(s, d)
expect(d.bucketCounts).to(equal(t[u'want']))
_expect_stats_eq_direct_calc_from_samples(d, samples)
def test_should_succeed_for_delta_metrics_with_the_distribution_type(self):
test_distribution = distribution.create_explicit([0.1, 0.3, 0.5])
distribution.add_sample(0.4, test_distribution)
v = metric_value.create(labels=self.TEST_LABELS,
distributionValue=test_distribution)
want = 2 * test_distribution.count
got = metric_value.merge(MetricKind.DELTA, v, v)
expect(got.distributionValue.count).to(equal(want))
def test_should_succeed_for_delta_metrics_with_the_distribution_type(self):
test_distribution = distribution.create_explicit([0.1, 0.3, 0.5])
distribution.add_sample(0.4, test_distribution)
v = metric_value.create(labels=self.TEST_LABELS,
distributionValue=test_distribution)
want = 2 * test_distribution.count
got = metric_value.merge(MetricKind.DELTA, v, v)
expect(got.distributionValue.count).to(equal(want))
def setUp(self):
self.merge_triples = ((
distribution.create_exponential(3, 2, 0.1),
distribution.create_exponential(3, 2, 0.1),
distribution.create_exponential(4, 2, 0.1),
), (distribution.create_linear(3, 0.2, 0.1),
distribution.create_linear(3, 0.2, 0.1),
distribution.create_linear(4, 0.2, 0.1)), (
distribution.create_explicit([0.1, 0.3]),
distribution.create_explicit([0.1, 0.3]),
distribution.create_explicit([0.1, 0.3, 0.5]),
))
for d1, d2, _ in self.merge_triples:
distribution.add_sample(_LOW_SAMPLE, d1)
distribution.add_sample(_HIGH_SAMPLE, d2)
def setUp(self):
self.merge_triples = (
(
distribution.create_exponential(3, 2, 0.1),
distribution.create_exponential(3, 2, 0.1),
distribution.create_exponential(4, 2, 0.1),
),(
distribution.create_linear(3, 0.2, 0.1),
distribution.create_linear(3, 0.2, 0.1),
distribution.create_linear(4, 0.2, 0.1)
),(
distribution.create_explicit([0.1, 0.3]),
distribution.create_explicit([0.1, 0.3]),
distribution.create_explicit([0.1, 0.3, 0.5]),
)
)
for d1, d2, _ in self.merge_triples:
distribution.add_sample(_LOW_SAMPLE, d1)
distribution.add_sample(_HIGH_SAMPLE, d2)
def test_should_fail_if_no_buckets_are_set(self):
testf = lambda: distribution.add_sample(_UNDERFLOW_SAMPLE,
self.NOTHING_SET)
expect(testf).to(raise_error(ValueError))
def _wanted_distribution_with_sample(value, *args):
d = distribution.create_exponential(*args)
distribution.add_sample(value, d)
return d
def _wanted_distribution_with_sample(value, *args):
d = distribution.create_exponential(*args)
distribution.add_sample(value, d)
return d
def test_should_fail_if_no_buckets_are_set(self):
testf = lambda: distribution.add_sample(_UNDERFLOW_SAMPLE, self.
NOTHING_SET)
expect(testf).to(raise_error(ValueError))