def test_STScan_subtask(model, project_task, analysis_model, grid_task):
provider = stscan.STScan(model)
subtask = standard_calls(provider, project_task, analysis_model, grid_task)
prediction = subtask(datetime.datetime(2017, 6, 7))
open_cp.stscan.STSTrainer.assert_called_with()
pred_mock = open_cp.stscan.STSTrainer.return_value
assert pred_mock.geographic_population_limit == 0.5
assert pred_mock.geographic_radius_limit == 3000
assert pred_mock.time_population_limit == 0.5
assert pred_mock.time_max_interval == np.timedelta64(60, "D")
np.testing.assert_array_equal(pred_mock.data.timestamps, [
np.datetime64("2017-05-21T12:30"),
np.datetime64("2017-05-21T13:00"),
np.datetime64("2017-05-21T13:30")
])
np.testing.assert_array_equal(pred_mock.data.xcoords, [0, 10, 20])
np.testing.assert_array_equal(pred_mock.data.ycoords, [10, 20, 0])
assert pred_mock.region.xmin == 25
assert pred_mock.region.ymin == 35
assert pred_mock.region.xmax == 25 + 50 * 15
assert pred_mock.region.ymax == 35 + 50 * 10
predict_date = np.datetime64(datetime.datetime(2017, 6, 7))
pred_mock.predict.assert_called_with(time=predict_date)
result = pred_mock.predict.return_value
result.grid_prediction.assert_called_with(50, use_maximal_clusters=False)
assert prediction is result.grid_prediction.return_value
def test_STScan_from_training_start(model, project_task, analysis_model,
grid_task):
analysis_model.time_range = (datetime.datetime(2017, 5, 21, 12,
59), None, None, None)
provider = stscan.STScan(model)
subtask = standard_calls(provider, project_task, analysis_model, grid_task)
prediction = subtask(datetime.datetime(2017, 6, 7))
pred_mock = open_cp.stscan.STSTrainer.return_value
np.testing.assert_array_equal(
pred_mock.data.timestamps,
[np.datetime64("2017-05-21T13:00"),
np.datetime64("2017-05-21T13:30")])
np.testing.assert_array_equal(pred_mock.data.xcoords, [10, 20])
np.testing.assert_array_equal(pred_mock.data.ycoords, [20, 0])
def test_STScan_max_clusters(model, project_task, analysis_model, grid_task):
provider = stscan.STScan(model)
provider.cluster_option = 2
assert provider.settings_string == "geo(50%/3000m) time(50%/60days) max"
data = provider.to_dict()
json_str = json.dumps(data)
provider.from_dict(json.loads(json_str))
assert provider.settings_string == "geo(50%/3000m) time(50%/60days) max"
subtask = standard_calls(provider, project_task, analysis_model, grid_task)
subtask(datetime.datetime(2017, 5, 21, 13, 0))
pred_mock = open_cp.stscan.STSTrainer.return_value
result = pred_mock.predict.return_value
result.grid_prediction.assert_called_with(50, use_maximal_clusters=True)
def test_STScan_time_window(model, project_task, analysis_model, grid_task):
provider = stscan.STScan(model)
provider.time_window_choice = 2
provider.time_window_length = datetime.timedelta(days=13)
assert provider.settings_string == "<=13days geo(50%/3000m) time(50%/60days)"
data = provider.to_dict()
json_str = json.dumps(data)
provider.from_dict(json.loads(json_str))
assert provider.settings_string == "<=13days geo(50%/3000m) time(50%/60days)"
subtask = standard_calls(provider, project_task, analysis_model, grid_task)
prediction = subtask(
datetime.datetime(2017, 5, 21, 13, 29) + datetime.timedelta(days=13))
pred_mock = open_cp.stscan.STSTrainer.return_value
np.testing.assert_array_equal(pred_mock.data.timestamps,
[np.datetime64("2017-05-21T13:30")])
np.testing.assert_array_equal(pred_mock.data.xcoords, [20])
np.testing.assert_array_equal(pred_mock.data.ycoords, [0])
def test_STScan_quant_both(model, project_task, analysis_model, grid_task):
provider = stscan.STScan(model)
provider.quantisation_choice = 4
provider.time_bin_length = datetime.timedelta(days=1)
assert provider.settings_string == "grid bins(24hours) geo(50%/3000m) time(50%/60days)"
data = provider.to_dict()
json_str = json.dumps(data)
provider.from_dict(json.loads(json_str))
assert provider.settings_string == "grid bins(24hours) geo(50%/3000m) time(50%/60days)"
subtask = standard_calls(provider, project_task, analysis_model, grid_task)
prediction = subtask(datetime.datetime(2017, 5, 21, 13, 0))
pred_mock = open_cp.stscan.STSTrainer.return_value
np.testing.assert_array_equal(pred_mock.data.timestamps, [
np.datetime64("2017-05-20T13:00"),
np.datetime64("2017-05-21T13:00"),
np.datetime64("2017-05-21T13:00")
])
np.testing.assert_array_equal(pred_mock.data.xcoords, [0, 0, 0])
np.testing.assert_array_equal(pred_mock.data.ycoords, [10, 10, 10])
def test_STScan_cluster_options(model, project_task, analysis_model,
grid_task):
provider = stscan.STScan(model)
provider.geographic_population_limit = 45
provider.time_population_limit = 55
provider.geographic_radius_limit = 1234
provider.time_max_interval = datetime.timedelta(days=23)
assert provider.settings_string == "geo(45%/1234m) time(55%/23days)"
data = provider.to_dict()
json_str = json.dumps(data)
provider.from_dict(json.loads(json_str))
assert provider.settings_string == "geo(45%/1234m) time(55%/23days)"
subtask = standard_calls(provider, project_task, analysis_model, grid_task)
prediction = subtask(datetime.datetime(2017, 6, 7))
pred_mock = open_cp.stscan.STSTrainer.return_value
assert pred_mock.geographic_population_limit == 0.45
assert pred_mock.geographic_radius_limit == 1234
assert pred_mock.time_population_limit == 0.55
assert pred_mock.time_max_interval == np.timedelta64(23, "D")
def test_STScan(model, project_task, analysis_model, grid_task):
provider = stscan.STScan(model)
assert provider.settings_string == "geo(50%/3000m) time(50%/60days)"
standard_calls(provider, project_task, analysis_model, grid_task)
def test_ProHotspot_serialise(model, project_task, analysis_model, grid_task):
serialise(stscan.STScan(model))