Feat/distribution showing

.github/workflows/basic-test-proxy.yaml

@@ -15,7 +15,8 @@ jobs:
       # setup repository with all necessary dependencies - keep as is
       - name: Check out repository
         uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - id: setup-python
+        uses: actions/setup-python@v4
         with:
           python-version: "3.11"
       - name: Install poetry

.github/workflows/run-bot-aib-tournament.yaml

@@ -20,7 +20,8 @@ jobs:
       - name: Check out repository
         uses: actions/checkout@v3
 
-      - uses: actions/setup-python@v4
+      - id: setup-python
+        uses: actions/setup-python@v4
         with:
           python-version: "3.11"
 

.github/workflows/run-bot-launcher.yaml

@@ -43,7 +43,8 @@ jobs:
       - name: Check out repository
         uses: actions/checkout@v3
 
-      - uses: actions/setup-python@v4
+      - id: setup-python
+        uses: actions/setup-python@v4
         with:
           python-version: "3.11"
 

.github/workflows/stress-test_proxy.yaml

@@ -14,7 +14,8 @@ jobs:
       # setup repository with all necessary dependencies - keep as is
       - name: Check out repository
         uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - id: setup-python
+        uses: actions/setup-python@v4
         with:
           python-version: "3.11"
       - name: Install poetry

code_tests/unit_tests/test_data_models/test_aggregate_edge_cases.py

@@ -0,0 +1,169 @@
+"""
+This  file is AI Generated. If it causes problems for a human we can probably just delete it.
+"""
+
+import numpy as np
+import pytest
+
+from forecasting_tools.data_models.numeric_report import (
+    NumericDistribution,
+    NumericReport,
+    Percentile,
+)
+from forecasting_tools.data_models.questions import NumericQuestion, QuestionState
+
+
+def _make_question(
+    lower_bound: float = 0.0,
+    upper_bound: float = 100e9,
+    open_lower_bound: bool = True,
+    open_upper_bound: bool = True,
+) -> NumericQuestion:
+    return NumericQuestion(
+        id_of_post=1,
+        id_of_question=1,
+        question_text="Test question",
+        background_info="",
+        resolution_criteria="",
+        fine_print="",
+        state=QuestionState.OPEN,
+        upper_bound=upper_bound,
+        lower_bound=lower_bound,
+        open_upper_bound=open_upper_bound,
+        open_lower_bound=open_lower_bound,
+        zero_point=None,
+    )
+
+
+def _make_distribution(
+    percentile_value_pairs: list[tuple[float, float]],
+    question: NumericQuestion,
+) -> NumericDistribution:
+    percentiles = [Percentile(percentile=p, value=v) for p, v in percentile_value_pairs]
+    return NumericDistribution.from_question(percentiles, question)
+
+
+CENTERED_DISTRIBUTION = [
+    (0.1, 30e9),
+    (0.2, 40e9),
+    (0.4, 50e9),
+    (0.6, 60e9),
+    (0.8, 70e9),
+    (0.9, 80e9),
+]
+SHIFTED_DISTRIBUTION = [
+    (0.1, 35e9),
+    (0.2, 45e9),
+    (0.4, 55e9),
+    (0.6, 65e9),
+    (0.8, 75e9),
+    (0.9, 85e9),
+]
+
+
+async def test_aggregate_shows_six_lines() -> None:
+    question = _make_question()
+    predictions = [
+        _make_distribution(CENTERED_DISTRIBUTION, question),
+        _make_distribution(SHIFTED_DISTRIBUTION, question),
+    ]
+    aggregated = await NumericReport.aggregate_predictions(predictions, question)
+    readable = NumericReport.make_readable_prediction(aggregated)
+    assert readable.count("chance of value below") == 6
+
+
+async def test_single_forecaster_aggregate() -> None:
+    question = _make_question()
+    predictions = [_make_distribution(CENTERED_DISTRIBUTION, question)]
+    aggregated = await NumericReport.aggregate_predictions(predictions, question)
+    readable = NumericReport.make_readable_prediction(aggregated)
+    assert "10.00%" in readable
+    assert "90.00%" in readable
+    result = aggregated.get_percentiles_at_target_heights()
+    assert result[0].value == pytest.approx(30e9, rel=0.1)
+    assert result[5].value == pytest.approx(80e9, rel=0.1)
+
+
+async def test_values_monotonically_increasing() -> None:
+    question = _make_question()
+    predictions = [
+        _make_distribution(CENTERED_DISTRIBUTION, question),
+        _make_distribution(SHIFTED_DISTRIBUTION, question),
+    ]
+    aggregated = await NumericReport.aggregate_predictions(predictions, question)
+    result = aggregated.get_percentiles_at_target_heights()
+    for i in range(len(result) - 1):
+        assert result[i].value < result[i + 1].value
+
+
+async def test_closed_bounds() -> None:
+    question = _make_question(open_lower_bound=False, open_upper_bound=False)
+    predictions = [
+        _make_distribution(CENTERED_DISTRIBUTION, question),
+        _make_distribution(SHIFTED_DISTRIBUTION, question),
+    ]
+    aggregated = await NumericReport.aggregate_predictions(predictions, question)
+    readable = NumericReport.make_readable_prediction(aggregated)
+    assert "10.00%" in readable
+
+
+async def test_nvidia_values_near_median() -> None:
+    question = _make_question()
+    forecaster_percentile_values = [
+        [(0.1, 55e9), (0.2, 58e9), (0.4, 61e9), (0.6, 63e9), (0.8, 66e9), (0.9, 70e9)],
+        [
+            (0.1, 54e9),
+            (0.2, 57e9),
+            (0.4, 60e9),
+            (0.6, 62.5e9),
+            (0.8, 65.5e9),
+            (0.9, 69e9),
+        ],
+        [
+            (0.1, 55e9),
+            (0.2, 58e9),
+            (0.4, 61e9),
+            (0.6, 62.5e9),
+            (0.8, 65e9),
+            (0.9, 68e9),
+        ],
+        [(0.1, 55e9), (0.2, 58e9), (0.4, 60e9), (0.6, 62e9), (0.8, 65e9), (0.9, 68e9)],
+    ]
+    predictions = [
+        _make_distribution(pvs, question) for pvs in forecaster_percentile_values
+    ]
+    aggregated = await NumericReport.aggregate_predictions(predictions, question)
+    result = aggregated.get_percentiles_at_target_heights()
+    for i in range(6):
+        expected = np.median([pvs[i][1] for pvs in forecaster_percentile_values])
+        assert result[i].value == pytest.approx(expected, rel=0.15)
+
+
+async def test_asymmetric_skew() -> None:
+    question = _make_question()
+    right_skewed_a = [
+        (0.1, 5e9),
+        (0.2, 8e9),
+        (0.4, 10e9),
+        (0.6, 12e9),
+        (0.8, 20e9),
+        (0.9, 50e9),
+    ]
+    right_skewed_b = [
+        (0.1, 6e9),
+        (0.2, 9e9),
+        (0.4, 11e9),
+        (0.6, 13e9),
+        (0.8, 22e9),
+        (0.9, 55e9),
+    ]
+    predictions = [
+        _make_distribution(right_skewed_a, question),
+        _make_distribution(right_skewed_b, question),
+    ]
+    aggregated = await NumericReport.aggregate_predictions(predictions, question)
+    readable = NumericReport.make_readable_prediction(aggregated)
+    assert "10.00%" in readable
+    result = aggregated.get_percentiles_at_target_heights()
+    for i in range(len(result) - 1):
+        assert result[i].value < result[i + 1].value