def test_flows_geojson_correct():
"""
Test that flows outputs expected geojson.
"""
dl1 = locate_subscribers("2016-01-01",
"2016-01-02",
level="admin3",
method="last")
dl2 = locate_subscribers("2016-01-02",
"2016-01-03",
level="admin3",
method="last")
flow = Flows(dl1, dl2)
fl_json = flow.to_geojson()
directory = os.path.dirname(os.path.os.path.realpath(__file__))
reference_file = os.path.join(directory, "./data/", "flows_reference.json")
with open(reference_file) as fin:
assert fl_json == json.load(fin)
def test_can_be_aggregated_admin3(get_dataframe):
"""
Query can be aggregated to a spatial level with admin3 data.
"""
mfl = locate_subscribers(
"2016-01-01", "2016-01-02", level="admin3", method="most-common"
)
agg = mfl.aggregate()
df = get_dataframe(agg)
assert ["pcod", "total"] == list(df.columns)
def test_can_be_joined(get_dataframe):
"""
RadiusOfGyration() can be joined with a location type metric.
"""
RoG = RadiusOfGyration("2016-01-01", "2016-01-02")
dl = locate_subscribers("2016-01-01", "2016-01-02", level="admin3")
rog_JA = RoG.join_aggregate(dl)
df = get_dataframe(rog_JA)
assert isinstance(df, pd.DataFrame)
assert rog_JA.column_names == ["name", "rog"]
def test_can_be_joined(get_dataframe):
"""
TopUpBalance() can be joined with a location type metric.
"""
topup_balance = TopUpBalance("2016-01-01", "2016-01-02", statistic="avg")
dl = locate_subscribers("2016-01-01",
"2016-01-02",
spatial_unit=make_spatial_unit("admin", level=3))
topup_balance_JA = topup_balance.join_aggregate(dl)
df = get_dataframe(topup_balance_JA)
assert topup_balance_JA.column_names == ["pcod", "value"]
def test_calculates_flows(get_dataframe):
"""
Flows() are correctly calculated
"""
dl1 = locate_subscribers("2016-01-01",
"2016-01-02",
level="admin3",
method="last")
dl2 = locate_subscribers("2016-01-02",
"2016-01-03",
level="admin3",
method="last")
flow = Flows(dl1, dl2)
df = get_dataframe(flow)
assert (df[(df.pcod_from == "524 3 09 50")
& (df.pcod_to == "524 5 14 73")]["count"].values[0] == 2)
assert (df[(df.pcod_from == "524 4 10 53")
& (df.pcod_to == "524 2 05 24")]["count"].values[0] == 2)
assert (df[(df.pcod_from == "524 1 02 09")
& (df.pcod_to == "524 3 08 44")]["count"].values[0] == 4)
def test_calculates_flows(get_dataframe):
"""
Flows() are correctly calculated
"""
dl1 = locate_subscribers("2016-01-01",
"2016-01-02",
level="admin3",
method="last")
dl2 = locate_subscribers("2016-01-02",
"2016-01-03",
level="admin3",
method="last")
flow = Flows(dl1, dl2)
df = get_dataframe(flow)
assert (df[(df.name_from == "Arghakhanchi")
& (df.name_to == "Dadeldhura")]["count"].values[0] == 2)
assert (df[(df.name_from == "Salyan")
& (df.name_to == "Kavrepalanchok")]["count"].values[0] == 2)
assert (df[(df.name_from == "Sankhuwasabha")
& (df.name_to == "Myagdi")]["count"].values[0] == 4)
def test_can_be_aggregated_admin3(self):
"""
Query can be aggregated to a spatial level with admin3 data.
"""
mfl = locate_subscribers("2016-01-01",
"2016-01-02",
level="admin3",
method="most-common")
agg = mfl.aggregate()
df = agg.get_dataframe()
self.assertIs(type(df), pd.DataFrame)
self.assertEqual(list(df.columns), ["name", "total"])
def test_most_fequent_admin(get_dataframe):
"""
Test that the most frequent admin3 is correctly calculated.
"""
mfl = locate_subscribers(
"2016-01-01", "2016-01-02", level="admin3", method="most-common"
)
df = get_dataframe(mfl)
# A few hand picked values
df_set = df.set_index("subscriber")["name"]
assert "Dolpa" == df_set["0gmvwzMAYbz5We1E"]
assert "Rukum" == df_set["1QBlwRo4Kd5v3Ogz"]
assert "Arghakhanchi" == df_set["2Dq97XmPqvL6noGk"]
def test_most_fequent_admin(get_dataframe):
"""
Test that the most frequent admin3 is correctly calculated.
"""
mfl = locate_subscribers(
"2016-01-01",
"2016-01-02",
spatial_unit=make_spatial_unit("admin", level=3),
method="most-common",
)
df = get_dataframe(mfl)
# A few hand picked values
df_set = df.set_index("subscriber")["pcod"]
assert "524 4 12 62" == df_set["0gmvwzMAYbz5We1E"]
assert "524 4 10 52" == df_set["1QBlwRo4Kd5v3Ogz"]
assert "524 3 09 50" == df_set["2Dq97XmPqvL6noGk"]
def test_can_be_aggregated_admin3_distribution(get_dataframe):
"""
Categorical queries can be aggregated to a spatial level with 'distribution' method.
"""
locations = locate_subscribers(
"2016-01-01",
"2016-01-02",
spatial_unit=make_spatial_unit("admin", level=3),
method="most-common",
)
metric = SubscriberHandsetCharacteristic("2016-01-01",
"2016-01-02",
characteristic="hnd_type")
agg = JoinedSpatialAggregate(metric=metric,
locations=locations,
method="distr")
df = get_dataframe(agg)
assert ["pcod", "metric", "key", "value"] == list(df.columns)
assert all(df[df.metric == "value"].groupby("pcod").sum() == 1.0)
