def test_pwo_result_storing():
"""Test that we can store a pwo model result and it will be retrieved for the next run"""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
r = p.run(departure_rate_vector={
"0xqNDj": 0.9
}, ignore_missing=True).store()
r.result()
r2 = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
assert r2.is_stored
def test_model_result_funcs():
"""Test overridden method for ModelResult"""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
mr = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
for ix, r in enumerate(mr):
pass
assert ix + 1 == len(mr)
mr2 = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
mr2.store().result()
for ix, r in enumerate(mr2):
pass
assert ix + 1 == len(mr)
def test_run_with_location_vector(get_dataframe):
"""
PopulationWeightedOpportunities().run() takes a location probability vector.
"""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-07")
set_df = get_dataframe(
p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True))
set_df = set_df.set_index("site_id_from")
assert set_df.loc["0xqNDj"]["site_id_to"].values[1] == "8wPojr"
assert set_df.loc["0xqNDj"]["site_id_to"].values[3] == "B8OaG5"
assert set_df.loc["0xqNDj"]["site_id_to"].values[7] == "DonxkP"
assert set_df.loc["0xqNDj"]["prediction"].values[1] == pytest.approx(
0.038338148548529)
assert set_df.loc["0xqNDj"]["prediction"].values[3] == pytest.approx(
0.143864222981437)
assert set_df.loc["0xqNDj"]["prediction"].values[7] == pytest.approx(
0.0618748599992082)
assert set_df.loc["0xqNDj"]["probability"].values[1] == pytest.approx(
0.00193627012871359)
assert set_df.loc["0xqNDj"]["probability"].values[3] == pytest.approx(
0.00726586984754731)
assert set_df.loc["0xqNDj"]["probability"].values[7] == pytest.approx(
0.00312499292925294)
def test_model_result_stored_len():
"""Test len of the stored model result is correct"""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
mr = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
l = len(mr)
mr.store().result()
assert l == len(mr)
def test_model_result_make_query():
"""Test that make query returns tablename and stores."""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
mr = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
qur = mr._make_query()
assert mr.is_stored
assert mr.table_name in qur
def test_model_result_cols():
"""Test model result object has right columns."""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
mr = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
cols = [
"site_id_from",
"version_from",
"lon_from",
"lat_from",
"site_id_to",
"version_to",
"lon_to",
"lat_to",
"prediction",
"probability",
]
assert mr.column_names == cols
mr.store().result()
assert mr.column_names == cols
def test_model_result_string_rep():
"""Test that ModelResult string rep has name of the model class"""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
mr = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
assert "PopulationWeightedOpportunities" in str(mr)
def test_get_stored():
"""Test that get_stored works for ModelResults."""
p = PopulationWeightedOpportunities("2016-01-01", "2016-01-02")
mr = p.run(departure_rate_vector={"0xqNDj": 0.9}, ignore_missing=True)
mr.store().result()
assert sum(1 for x in PopulationWeightedOpportunities.get_stored()) == 1