def setUp(self):
self.rog = RadiusOfGyration("2016-01-01", "2016-01-02")
self.low = 150
self.high = 155
self.rog_df = self.rog.get_dataframe().query(
"{low} <= rog <= {high}".format(low=self.low, high=self.high))
self.sub = self.rog.numeric_subset(col="rog",
low=self.low,
high=self.high)
def test_joined_agg_date_mismatch():
"""
Test that join aggregate with mismatched dates raises a warning.
"""
mfl = MostFrequentLocation("2016-01-01", "2016-01-04", level="admin3")
with pytest.warns(UserWarning):
mfl.join_aggregate(RadiusOfGyration("2016-01-02", "2016-01-04"))
with pytest.warns(UserWarning):
mfl.join_aggregate(RadiusOfGyration("2016-01-01", "2016-01-05"))
def test_can_numsubset_with_inf(get_dataframe):
"""
flowmachine.RadiusOfGyration can be subset between -Inf and Inf
"""
rog = RadiusOfGyration("2016-01-01", "2016-01-02")
low = -float("Infinity")
high = float("Infinity")
sub = get_dataframe(rog.numeric_subset(col="value", low=low, high=high))
df = get_dataframe(rog).query("{low} <= value <= {high}".format(low=low,
high=high))
pd.testing.assert_frame_equal(sub, df)
def test_joined_median_aggregate(self):
"""
Test join with median aggregate.
"""
mfl = MostFrequentLocation("2016-01-01", "2016-01-04", level="admin3")
rog = RadiusOfGyration("2016-01-01", "2016-01-04")
joined = mfl.join_aggregate(rog, method="median")
rawus_avg = (rog.get_dataframe().set_index("subscriber").join(
mfl.get_dataframe().set_index("subscriber")).set_index(
"name").ix["Rasuwa"].rog.median())
self.assertAlmostEqual(
joined.get_dataframe().set_index("name").ix["Rasuwa"].rog,
rawus_avg)
def test_can_numsubset_with_low_and_high(get_dataframe):
"""
flowmachine.RadiusOfGyration can be subset within a range
"""
rog = RadiusOfGyration("2016-01-01", "2016-01-02")
low = 150
high = 155
rog_df = (get_dataframe(rog).query("{low} <= value <= {high}".format(
low=low, high=high)).set_index("subscriber"))
sub = get_dataframe(rog.numeric_subset(col="value", low=low,
high=high)).set_index("subscriber")
pd.testing.assert_frame_equal(sub, rog_df)
def test_num_subset_can_be_stored(get_dataframe):
"""
Test that flowmachine.NumericSubset can be stored.
"""
rog = RadiusOfGyration("2016-01-01", "2016-01-02")
low = 150
high = 155
rog_df = get_dataframe(rog).query("{low} <= value <= {high}".format(
low=low, high=high))
sub = rog.numeric_subset(col="value", low=low, high=high)
sub.store().result()
assert sub.is_stored
# Test that the store is of the right length
sub = rog.numeric_subset(col="value", low=low, high=high)
assert len(get_dataframe(sub)) == len(rog_df)
def test_dropna(get_length):
"""
Test that we are able to keep rows with NA values.
"""
start, stop = "2016-01-01", "2016-01-03"
msisdn = "1vGR8kp342yxEpwY"
sql = """
select
msisdn as subscriber,
2 as val
from
events.calls
where
msisdn = '{}'
limit 1
""".format(
msisdn
)
metrics = [CustomQuery(sql, ["subscriber"]), RadiusOfGyration(start, stop)]
fc = feature_collection(metrics, dropna=False)
# usully without dropna=False this query would only return
# a single row. We check that this is not the case.
assert get_length(fc) > 1
def test_joined_agg_hours_mismatch():
"""
Test that join aggregate with mismatched hours doesn't warn.
"""
mfl = MostFrequentLocation("2016-01-01 10:00", "2016-01-04", level="admin3")
with warnings.catch_warnings(record=True) as w:
mfl.join_aggregate(RadiusOfGyration("2016-01-01", "2016-01-04"))
assert not w
def test_histogram_param_value_errors(param_name, param_value,
expected_exception):
radius_of_gyration = RadiusOfGyration("2016-01-01", "2016-01-02")
args = dict(bins=10, metric=radius_of_gyration)
args[param_name] = param_value
with pytest.raises(ValueError, match=expected_exception):
HistogramAggregation(**args)
def test_can_get_item_subscriber_metric(get_dataframe):
"""g
flowmachine.SubscriberFeature allows for getting items
"""
rog = RadiusOfGyration("2016-01-01", "2016-01-03")
dl = daily_location("2016-01-03")
single_subscriber = list(get_dataframe(dl).head(8).subscriber)[3]
sub = get_dataframe(rog[single_subscriber])
assert set(sub.subscriber) == {single_subscriber}
def test_joined_aggregate(get_dataframe):
"""
Test join aggregate.
"""
mfl = MostFrequentLocation("2016-01-01", "2016-01-04", level="admin3")
joined = mfl.join_aggregate(RadiusOfGyration("2016-01-01", "2016-01-04"))
assert (
pytest.approx(203.12391560786)
== get_dataframe(joined).set_index("pcod").loc["524 2 05 29"].rog
)
def test_joined_aggregate(get_dataframe):
"""
Test join aggregate.
"""
mfl = MostFrequentLocation("2016-01-01", "2016-01-04", level="admin3")
joined = mfl.join_aggregate(RadiusOfGyration("2016-01-01", "2016-01-04"))
assert (
pytest.approx(199.956021886114)
== get_dataframe(joined).set_index("name").ix["Rasuwa"].rog
)
def test_joined_aggregate(self):
"""
Test join aggregate.
"""
mfl = MostFrequentLocation("2016-01-01", "2016-01-04", level="admin3")
joined = mfl.join_aggregate(
RadiusOfGyration("2016-01-01", "2016-01-04"))
self.assertAlmostEqual(
joined.get_dataframe().set_index("name").ix["Rasuwa"].rog,
199.956021886114)
def _flowmachine_query_obj(self):
"""
Return the underlying flowmachine radius_of_gyration object.
Returns
-------
Query
"""
return RadiusOfGyration(
start=self.start_date,
stop=self.end_date,
subscriber_subset=self.subscriber_subset,
)
def test_create_histogram_using_int_bins_value(get_dataframe):
"""
Create histogram using one bins value.
"""
radius_of_gyration = RadiusOfGyration("2016-01-01", "2016-01-02")
agg = HistogramAggregation(metric=radius_of_gyration, bins=5, censor=False)
df = get_dataframe(agg)
numpy_histogram, numpy_bins = np.histogram(
get_dataframe(radius_of_gyration).value, bins=5)
assert df.value.sum() == len(get_dataframe(radius_of_gyration))
assert numpy_histogram.tolist() == df.value.tolist()
assert numpy_bins.tolist()[:-1] == pytest.approx(df.lower_edge.tolist())
assert numpy_bins.tolist()[1:] == pytest.approx(df.upper_edge.tolist())
def test_query_can_be_subscriber_set_restricted(self):
"""Test that some queries can be limited to only a subset of subscribers."""
# Create a temporary table in the DB
con = Table.connection.engine
sql = "DROP TABLE IF EXISTS subscriber_list"
con.execute(sql)
sql = """CREATE TABLE subscriber_list (subscriber TEXT)"""
con.execute(sql)
formatted_subscribers = ",".join("('{}')".format(u)
for u in self.subscriber_list)
sql = """INSERT INTO subscriber_list (subscriber) VALUES {}""".format(
formatted_subscribers)
con.execute(sql)
rog = RadiusOfGyration("2016-01-01",
"2016-01-03",
subscriber_subset=Table("subscriber_list"))
hl = HomeLocation(*[
daily_location(d, subscriber_subset=Table("subscriber_list"))
for d in list_of_dates("2016-01-01", "2016-01-03")
])
rog_df = rog.get_dataframe()
hl_df = hl.get_dataframe()
sql = "DROP TABLE IF EXISTS subscriber_list"
con.execute(sql)
# Get the set of subscribers present in the dataframe, we need to handle the logic
# of msisdn_from/msisdn_to
calculated_subscriber_set = set(rog_df.subscriber)
self.assertEqual(calculated_subscriber_set, set(self.subscriber_list))
calculated_subscriber_set = set(hl_df.subscriber)
self.assertEqual(calculated_subscriber_set, set(self.subscriber_list))
def _unsampled_query_obj(self):
"""
Return the underlying flowmachine radius_of_gyration object.
Returns
-------
Query
"""
return RadiusOfGyration(
start=self.start_date,
stop=self.end_date,
table=self.event_types,
subscriber_subset=self.subscriber_subset,
hours=self.hours,
)
def test_create_histogram_using_list_of_bins_values(get_dataframe):
"""
Create histogram using list of bins values.
"""
radius_of_gyration = RadiusOfGyration("2016-01-01", "2016-01-02")
agg = HistogramAggregation(metric=radius_of_gyration,
bins=[10, 20, 30, 40, 50, 60],
censor=False)
df = get_dataframe(agg)
numpy_histogram, numpy_bins = np.histogram(
get_dataframe(radius_of_gyration).value, bins=[10, 20, 30, 40, 50, 60])
assert numpy_histogram.tolist() == df.value.tolist()
assert numpy_bins.tolist()[:-1] == pytest.approx(df.lower_edge.tolist())
assert numpy_bins.tolist()[1:] == pytest.approx(df.upper_edge.tolist())
def test_call_with_str_raises_error():
"""
Numeric subset can't be called with a string in arguments low and high
"""
rog = RadiusOfGyration("2016-01-01", "2016-01-02")
with pytest.raises(TypeError):
rog.numeric_subset(col="value", low="foo", high=1)
with pytest.raises(TypeError):
rog.numeric_subset(col="value", low=1, high="bar")
def test_all_above_threshold(get_dataframe):
"""
Test that values are not returned where there are not enough people in the aggregate.
"""
in_agg = get_dataframe(
RedactedJoinedSpatialAggregate(
joined_spatial_aggregate=JoinedSpatialAggregate(
locations=daily_location("2016-01-01"),
metric=RadiusOfGyration("2016-01-01", "2016-01-02"),
))).pcod
assert len(in_agg) > 0
under_15 = get_dataframe(
daily_location("2016-01-01").aggregate().numeric_subset(col="value",
low=0,
high=15)).pcod
assert set(under_15).isdisjoint(in_agg)
def test_create_histogram_using_bins_and_range_values(get_dataframe):
"""
Create histogram using one bins and range values.
"""
radius_of_gyration = RadiusOfGyration("2016-01-01", "2016-01-02")
agg = HistogramAggregation(metric=radius_of_gyration,
bins=5,
range=(130.00, 230.00),
censor=False)
df = get_dataframe(agg)
numpy_histogram, numpy_bins = np.histogram(
get_dataframe(radius_of_gyration).value,
bins=5,
range=(130.00, 230.00))
assert numpy_histogram.tolist() == df.value.tolist()
assert numpy_bins.tolist()[:-1] == pytest.approx(df.lower_edge.tolist())
assert numpy_bins.tolist()[1:] == pytest.approx(df.upper_edge.tolist())
def test_joined_median_aggregate(get_dataframe):
"""
Test join with median aggregate.
"""
mfl = MostFrequentLocation("2016-01-01", "2016-01-04", level="admin3")
rog = RadiusOfGyration("2016-01-01", "2016-01-04")
joined = mfl.join_aggregate(rog, method="median")
rawus_avg = (
get_dataframe(rog)
.set_index("subscriber")
.join(get_dataframe(mfl).set_index("subscriber"))
.set_index("pcod")
.loc["524 2 05 29"]
.rog.median()
)
assert (
pytest.approx(rawus_avg)
== get_dataframe(joined).set_index("pcod").loc["524 2 05 29"].rog
), rawus_avg
def test_collects_metrics():
"""
Test that we can instantiate flowmachine.feature_collection with list of
objects.
"""
start, stop = "2016-01-01", "2016-01-03"
metrics = [
RadiusOfGyration(start, stop),
NocturnalEvents(start, stop),
SubscriberDegree(start, stop),
]
expected_columns = [
"subscriber",
"value_radiusofgyration_0",
"value_nocturnalevents_1",
"value_subscriberdegree_2",
]
fc = feature_collection(metrics)
column_names = fc.column_names
assert expected_columns == column_names
def test_create_histogram_using_bins_list_and_range_values(get_dataframe):
"""
Create histogram using list of bins and range values (checking for consistency with numpy).
"""
radius_of_gyration = RadiusOfGyration("2016-01-01", "2016-01-02")
agg = HistogramAggregation(
metric=radius_of_gyration,
bins=[10, 20, 30, 40, 50, 60],
range=(130.00, 230.00),
censor=False,
)
df = get_dataframe(agg)
numpy_histogram, numpy_bins = np.histogram(
get_dataframe(radius_of_gyration).value,
bins=[10, 20, 30, 40, 50, 60],
range=(130.00, 230.00),
)
assert numpy_histogram.tolist() == df.value.tolist()
assert numpy_bins.tolist()[:-1] == pytest.approx(df.lower_edge.tolist())
assert numpy_bins.tolist()[1:] == pytest.approx(df.upper_edge.tolist())
def test_query_can_be_subscriber_set_restricted(subscriber_list_table,
subscriber_list,
get_dataframe):
"""Test that some queries can be limited to only a subset of subscribers."""
rog = RadiusOfGyration("2016-01-01",
"2016-01-03",
subscriber_subset=subscriber_list_table)
hl = ModalLocation(*[
daily_location(d, subscriber_subset=subscriber_list_table)
for d in list_of_dates("2016-01-01", "2016-01-03")
])
rog_df = get_dataframe(rog)
hl_df = get_dataframe(hl)
# Get the set of subscribers present in the dataframe, we need to handle the logic
# of msisdn_from/msisdn_to
calculated_subscriber_set = set(rog_df.subscriber)
assert calculated_subscriber_set == set(subscriber_list)
calculated_subscriber_set = set(hl_df.subscriber)
assert calculated_subscriber_set == set(subscriber_list)
def setUp(self):
self.dl = daily_location("2016-01-03")
self.rog = RadiusOfGyration("2016-01-01", "2016-01-03")
self.subscriber_list = list(self.dl.head(8).subscriber)
self.single_subscriber = self.subscriber_list[3]
class TestNumericSubsetting(TestCase):
def setUp(self):
self.rog = RadiusOfGyration("2016-01-01", "2016-01-02")
self.low = 150
self.high = 155
self.rog_df = self.rog.get_dataframe().query(
"{low} <= rog <= {high}".format(low=self.low, high=self.high))
self.sub = self.rog.numeric_subset(col="rog",
low=self.low,
high=self.high)
def _query_has_values(self, Q, expected_df):
"""
Test if the values of a dataframes columns are equal
to certain values.
"""
query_df = Q.get_dataframe()
assert_array_equal(expected_df.values, query_df.values)
def test_can_numsubset_with_low_and_high(self):
"""
flowmachine.RadiusOfGyration can be subset within a range
"""
self._query_has_values(self.sub, self.rog_df)
def test_can_numsubset_with_inf(self):
"""
flowmachine.RadiusOfGyration can be subset between -Inf and Inf
"""
low = -float("Infinity")
high = float("Infinity")
sub = self.rog.numeric_subset(col="rog", low=low, high=high)
df = self.rog.get_dataframe().query("{low} <= rog <= {high}".format(
low=low, high=high))
self._query_has_values(sub, df)
def test_call_with_str_raises_error(self):
"""
Numeric subset can't be called with a string in arguments low and high
"""
with self.assertRaises(TypeError):
self.rog.numeric_subset(col="rog", low="foo", high=self.high)
with self.assertRaises(TypeError):
self.rog.numeric_subset(col="rog", low=self.low, high="bar")
def test_can_be_stored(self):
"""
Test that flowmachine.NumericSubset can be stored.
"""
self.sub.store().result()
self.assertTrue(self.sub.is_stored)
# Test that the store is of the right length
sub = self.rog.numeric_subset(col="rog", low=self.low, high=self.high)
self.assertEqual(len(sub.get_dataframe()), len(self.rog_df))
def tearDown(self):
"""
Remove stored table from "can_be_stored" test.
"""
self.sub.invalidate_db_cache()
