def test_base_cascade_model_obs_irrelevant():
np.random.seed(42)
y = [1, 2, 0, 4, 3]
assert click(BaseCascadeModel(0.3, 0), [], y) == [1, 1, 1, 1, 1]
assert click(BaseCascadeModel(0.3, 1), [], y) == [1, 1, 0, 1, 0]
assert click(BaseCascadeModel(0.3, 3), [], y) == [0, 0, 0, 1, 1]
assert click(BaseCascadeModel(0.3, 4), [], y) == [0, 0, 0, 1, 0]
def test_only_relevant_above_threshold_click_model():
click_model = OnlyRelevantClickModel(2)
assert click(click_model, np.array([[0, 1]]), [2]) == [1]
assert click(click_model, np.array([[0, 1]]), [1]) == [0]
assert click(click_model, np.array([[0, 1]]), [0]) == [0]
assert click(click_model, np.array([[1, 1], [1, 0], [0, 0]]),
[0, 1, 2]) == [0, 0, 1]
def test_inner_click_model_should_just_get_unmasked_docs():
np.random.seed(42)
click_model = MaskedRemainMasked(RandomClickModel(n_clicks=1))
clicks = click(
click_model, np.ones((5, 1)),
np.array([
0, PADDED_Y_VALUE, PADDED_Y_VALUE, PADDED_Y_VALUE, PADDED_Y_VALUE
]))
assert clicks == [
1, PADDED_Y_VALUE, PADDED_Y_VALUE, PADDED_Y_VALUE, PADDED_Y_VALUE
]
def test_base_cascade_model_eta():
np.random.seed(42)
click_model_1 = BaseCascadeModel(0.3, 1)
click_model_2 = BaseCascadeModel(0.5, 1)
assert click(click_model_1, [], [1, 2]) == [1, 0]
assert click(click_model_1, [], [1, 2, 3]) == [1, 1, 1]
assert click(click_model_1, [], [1, 2, 3, 4]) == [1, 1, 0, 1]
assert click(click_model_2, [], [1, 2]) == [1, 1]
assert click(click_model_2, [], [1, 2, 3]) == [1, 0, 1]
assert click(click_model_2, [], [1, 2, 3, 4]) == [1, 1, 1, 0]
def test_masked_should_remain_masked():
click_model = MaskedRemainMasked(FixedClickModel(click_positions=[1]))
assert click(click_model, np.ones((3, 1)),
np.array([0, 0, PADDED_Y_VALUE])) == [0, 1, PADDED_Y_VALUE]
def test_fixed_click_model_single():
click_model = FixedClickModel([0])
assert click(click_model, [], [1]) == [1]
assert click(click_model, [], [1, 2]) == [1, 0]
assert click(click_model, [], [1, 2, 3]) == [1, 0, 0]
def test_fixed_click_model_multiple():
assert click(FixedClickModel([0, 1]), [], [1, 2, 3, 4]) == [1, 1, 0, 0]
assert click(FixedClickModel([0, 1, 2]), [], [1, 2, 3, 4]) == [1, 1, 1, 0]
assert click(FixedClickModel([0, 2, 3]), [], [1, 2, 3, 4]) == [1, 0, 1, 1]
def test_feature_click_model_everything():
click_model = EverythingButDuplicatesClickModel()
assert click(click_model, np.array([[0, 1]]), []) == [1]
assert click(click_model, np.array([[1, 1], [1, 0]]), []) == [1, 1]
assert click(click_model, np.array([[1, 1], [1, 0], [0, 0]]),
[]) == [1, 1, 1]
def test_feature_click_model_except_near_duplicates():
click_model = EverythingButDuplicatesClickModel(0.1)
assert click(click_model, np.array([[0, 1]]), []) == [1]
assert click(click_model, np.array([[1, 1], [1, 1]]), []) == [1, 0]
assert click(click_model, np.array([[1, 1], [1, 0.99], [1, 0.8]]),
[]) == [1, 0, 1]
def test_only_relevant_click_model():
click_model = OnlyRelevantClickModel(1)
assert click(click_model, np.array([[0, 1]]), [1]) == [1]
assert click(click_model, np.array([[0, 1]]), [0]) == [0]
assert click(click_model, np.array([[1, 1], [1, 0], [0, 0]]),
[1, 0, 0]) == [1, 0, 0]
def test_random_click_model_single():
click_model = RandomClickModel(1)
np.random.seed(42)
assert click(click_model, [], [1]) == [1]
assert click(click_model, [], [1, 2]) == [0, 1]
assert click(click_model, [], [1, 2, 3]) == [0, 1, 0]
def test_random_click_model_multiple():
np.random.seed(42)
assert click(RandomClickModel(2), [], [1, 2, 3, 4]) == [0, 1, 0, 1]
assert click(RandomClickModel(3), [], [1, 2, 3, 4]) == [1, 1, 0, 1]
assert click(RandomClickModel(4), [], [1, 2, 3, 4]) == [1, 1, 1, 1]
def test_base_cascade_model_no_eta():
click_model = BaseCascadeModel(0.0, 1)
assert click(click_model, [], [1]) == [1]
assert click(click_model, [], [1, 2]) == [1, 1]
assert click(click_model, [], [1, 2, 3]) == [1, 1, 1]
def test_base_cascade_model_below_threshold():
y = [1, 2, 0, 4, 3]
assert click(BaseCascadeModel(0.0, 1), [], y) == [1, 1, 0, 1, 1]
assert click(BaseCascadeModel(0.0, 2), [], y) == [0, 1, 0, 1, 1]
assert click(BaseCascadeModel(0.0, 4), [], y) == [0, 0, 0, 1, 0]
