Skip to content

mc/3806/backend/aggregations #3818

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
lsabor wants to merge 2 commits into
base: mc/3813/backend/forecasts
Choose a base branch
from
Open

mc/3806/backend/aggregations #3818

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6a938a3
add options_history field to question model and migration
lsabor Nov 15, 2025
c70aff2
mc/3806/aggregations
lsabor Nov 20, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions tests/unit/test_questions/conftest.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@

__all__ = [
"question_binary",
"question_multiple_choice",
"question_numeric",
"conditional_1",
"question_binary_with_forecast_user_1",
Expand Down
5 changes: 4 additions & 1 deletion tests/unit/test_utils/test_the_math/conftest.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1 +1,4 @@
from tests.unit.test_questions.conftest import question_binary # noqa
from tests.unit.test_questions.conftest import (
question_binary,
question_multiple_choice,
) # noqa
228 changes: 199 additions & 29 deletions tests/unit/test_utils/test_the_math/test_aggregations.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,12 @@
GoldMedalistsAggregation,
JoinedBeforeDateAggregation,
SingleAggregation,
compute_weighted_semi_standard_deviations,
)
from utils.typing import (
ForecastValues,
ForecastsValues,
Weights,
)


Expand All @@ -46,6 +52,64 @@ def test_summarize_array(array, max_size, expceted_array):

class TestAggregations:

@pytest.mark.parametrize(
"forecasts_values, weights, expected",
[
(
[[0.5, 0.5]],
None,
([0.0, 0.0], [0.0, 0.0]),
), # Trivial
(
[
[0.5, 0.5],
[0.5, 0.5],
[0.5, 0.5],
],
None,
([0.0, 0.0], [0.0, 0.0]),
), # 3 unwavaring forecasts
(
[
[0.2, 0.8],
[0.5, 0.5],
[0.8, 0.2],
],
None,
([0.3, 0.3], [0.3, 0.3]),
), # 3 unwavaring forecasts
(
[
[0.6, 0.15, None, 0.25],
[0.6, 0.15, None, 0.25],
],
None,
([0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0]),
), # identical forecasts with placeholders
(
[
[0.4, 0.25, None, 0.35],
[0.6, 0.15, None, 0.25],
],
None,
([0.1, 0.05, 0.0, 0.05], [0.1, 0.05, 0.0, 0.05]),
), # minorly different forecasts with placeholders
],
)
def test_compute_weighted_semi_standard_deviations(
self,
forecasts_values: ForecastsValues,
weights: Weights | None,
expected: tuple[ForecastValues, ForecastValues],
):
result = compute_weighted_semi_standard_deviations(forecasts_values, weights)
rl, ru = result
el, eu = expected
for v, e in zip(rl, el):
np.testing.assert_approx_equal(v, e)
for v, e in zip(ru, eu):
np.testing.assert_approx_equal(v, e)

@pytest.mark.parametrize("aggregation_name", [Agg.method for Agg in AGGREGATIONS])
def test_aggregations_initialize(
self, question_binary: Question, aggregation_name: str
Expand Down Expand Up @@ -241,46 +305,120 @@ def test_aggregations_initialize(
histogram=None,
),
),
# Multiple choice with placeholders
(
{},
ForecastSet(
forecasts_values=[
[0.6, 0.15, None, 0.25],
[0.6, 0.25, None, 0.15],
],
timestep=datetime(2024, 1, 1, tzinfo=dt_timezone.utc),
forecaster_ids=[1, 2],
timesteps=[
datetime(2022, 1, 1, tzinfo=dt_timezone.utc),
datetime(2023, 1, 1, tzinfo=dt_timezone.utc),
],
),
True,
False,
AggregateForecast(
start_time=datetime(2024, 1, 1, tzinfo=dt_timezone.utc),
method=AggregationMethod.UNWEIGHTED,
forecast_values=[0.6, 0.20, None, 0.20],
interval_lower_bounds=[0.6, 0.15, None, 0.15],
centers=[0.6, 0.20, None, 0.20],
interval_upper_bounds=[0.6, 0.25, None, 0.25],
means=[0.6, 0.20, None, 0.20],
forecaster_count=2,
),
),
(
{},
ForecastSet(
forecasts_values=[
[0.6, 0.15, None, 0.25],
[0.6, 0.25, None, 0.15],
[0.4, 0.35, None, 0.25],
],
timestep=datetime(2024, 1, 1, tzinfo=dt_timezone.utc),
forecaster_ids=[1, 2],
timesteps=[
datetime(2022, 1, 1, tzinfo=dt_timezone.utc),
datetime(2023, 1, 1, tzinfo=dt_timezone.utc),
],
),
True,
False,
AggregateForecast(
start_time=datetime(2024, 1, 1, tzinfo=dt_timezone.utc),
method=AggregationMethod.UNWEIGHTED,
forecast_values=[
0.5453965360072925,
0.22730173199635367,
None,
0.22730173199635367,
],
interval_lower_bounds=[
0.3635976906715284,
0.1363810391978122,
None,
0.1363810391978122,
],
centers=[
0.5453965360072925,
0.22730173199635367,
None,
0.22730173199635367,
],
interval_upper_bounds=[
0.5453965360072925,
0.3182224247948951,
None,
0.22730173199635367,
],
means=[
0.5333333333333333,
0.25,
None,
0.21666666666666667,
],
forecaster_count=3,
),
),
],
)
def test_UnweightedAggregation(
self,
question_binary: Question,
question_multiple_choice: Question,
init_params: dict,
forecast_set: ForecastSet,
include_stats: bool,
histogram: bool,
expected: AggregateForecast,
):
aggregation = UnweightedAggregation(question=question_binary, **init_params)
new_aggregation = aggregation.calculate_aggregation_entry(
if len(forecast_set.forecasts_values[0]) == 2:
question = question_binary
else:
question = question_multiple_choice

aggregation = UnweightedAggregation(question=question, **init_params)
new_aggregation: AggregateForecast = aggregation.calculate_aggregation_entry(
forecast_set, include_stats, histogram
)

assert new_aggregation.start_time == expected.start_time
assert (
new_aggregation.forecast_values == expected.forecast_values
) or np.allclose(new_aggregation.forecast_values, expected.forecast_values)
assert new_aggregation.forecaster_count == expected.forecaster_count
assert (
new_aggregation.interval_lower_bounds == expected.interval_lower_bounds
) or np.allclose(
new_aggregation.interval_lower_bounds, expected.interval_lower_bounds
)
assert (new_aggregation.centers == expected.centers) or np.allclose(
new_aggregation.centers, expected.centers
)
assert (
new_aggregation.interval_upper_bounds == expected.interval_upper_bounds
) or np.allclose(
new_aggregation.interval_upper_bounds, expected.interval_upper_bounds
)
assert (new_aggregation.means == expected.means) or np.allclose(
new_aggregation.means, expected.means
)
assert (new_aggregation.histogram == expected.histogram) or np.allclose(
new_aggregation.histogram, expected.histogram
)
for r, e in [
(new_aggregation.forecast_values, expected.forecast_values),
(new_aggregation.interval_lower_bounds, expected.interval_lower_bounds),
(new_aggregation.centers, expected.centers),
(new_aggregation.interval_upper_bounds, expected.interval_upper_bounds),
(new_aggregation.means, expected.means),
(new_aggregation.histogram, expected.histogram),
]:
r = np.where(np.equal(r, None), np.nan, r).astype(float)
e = np.where(np.equal(e, None), np.nan, e).astype(float)
np.testing.assert_allclose(r, e, equal_nan=True)

@pytest.mark.parametrize(
"init_params, forecast_set, include_stats, histogram, expected",
Expand Down Expand Up @@ -468,20 +606,52 @@ def test_UnweightedAggregation(
histogram=None,
),
),
# Multiple choice with placeholders
(
{},
ForecastSet(
forecasts_values=[
[0.6, 0.15, None, 0.25],
[0.6, 0.25, None, 0.15],
],
timestep=datetime(2024, 1, 1, tzinfo=dt_timezone.utc),
forecaster_ids=[1, 2],
timesteps=[
datetime(2022, 1, 1, tzinfo=dt_timezone.utc),
datetime(2023, 1, 1, tzinfo=dt_timezone.utc),
],
),
True,
False,
AggregateForecast(
start_time=datetime(2024, 1, 1, tzinfo=dt_timezone.utc),
method=AggregationMethod.UNWEIGHTED,
forecast_values=[0.6, 0.20, None, 0.20],
interval_lower_bounds=[0.6, 0.15, None, 0.15],
centers=[0.6, 0.20, None, 0.20],
interval_upper_bounds=[0.6, 0.25, None, 0.25],
means=[0.6, 0.20, None, 0.20],
forecaster_count=2,
),
),
],
)
def test_RecencyWeightedAggregation(
self,
question_binary: Question,
question_multiple_choice: Question,
init_params: dict,
forecast_set: ForecastSet,
include_stats: bool,
histogram: bool,
expected: AggregateForecast,
):
aggregation = RecencyWeightedAggregation(
question=question_binary, **init_params
)
if len(forecast_set.forecasts_values[0]) == 2:
question = question_binary
else:
question = question_multiple_choice

aggregation = RecencyWeightedAggregation(question=question, **init_params)
new_aggregation = aggregation.calculate_aggregation_entry(
forecast_set, include_stats, histogram
)
Expand Down
Loading