Derive default end year from CPI-U YAML and reorganize test fixtures

- Replace hardcoded DEFAULT_END_YEAR = 2035 with dynamic derivation
  from the CPI-U parameter YAML (gov.bls.cpi.cpi_u), so datasets
  automatically extend to whatever year the CPI-U is updated to
- Add get_parameter_last_year() utility that reads a parameter's
  source YAML, sorts date keys, and returns the latest year (reads
  YAML on disk, not runtime values_list, to avoid the 2100
  extrapolation from uprating_extensions.py)
- Rename fixtures/economic_assumptions_fixtures.py to
  fixtures/test_extend_single_year_dataset.py to match test file name
- Move all test setup code (helpers, mocks, pytest fixtures) from
  test file into fixtures, leaving only test classes in test file

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: anth-volk

policyengine_us/data/__init__.py

@@ -1,2 +1,2 @@
 from .dataset_schema import USSingleYearDataset, USMultiYearDataset
-from .economic_assumptions import extend_single_year_dataset
+from .economic_assumptions import extend_single_year_dataset, get_parameter_last_year

policyengine_us/data/economic_assumptions.py

@@ -1,14 +1,44 @@
+import yaml
+
 from policyengine_us.data.dataset_schema import (
     USSingleYearDataset,
     USMultiYearDataset,
 )
 
-DEFAULT_END_YEAR = 2035
+# The default end year for dataset extension is derived at runtime from
+# the CPI-U parameter YAML (gov.bls.cpi.cpi_u).  When the CPI-U YAML
+# is updated with new projection years, datasets will automatically
+# extend to match — no hardcoded year constant to maintain.
+CPI_U_PARAM_PATH = "gov.bls.cpi.cpi_u"
+
+
+def get_parameter_last_year(parameter) -> int:
+    """Return the latest year explicitly defined in a parameter's YAML file.
+
+    Reads the YAML source file (via ``parameter.file_path``), parses
+    the ``values`` mapping, sorts its date keys chronologically, and
+    returns the year component of the latest entry.
+
+    This deliberately reads the YAML on disk rather than the runtime
+    ``values_list``, because ``uprating_extensions.py`` programmatically
+    extends many parameters to 2100 after loading — and we want the
+    last *authored* year, not the extrapolated one.
+    """
+    with open(parameter.file_path) as f:
+        data = yaml.safe_load(f)
+    date_keys = sorted(str(k) for k in data.get("values", {}).keys())
+    return int(date_keys[-1][:4])
+
+
+def _get_default_end_year(system) -> int:
+    """Derive the default end year from the CPI-U parameter's YAML."""
+    cpi_u = _resolve_parameter(system.parameters, CPI_U_PARAM_PATH)
+    return get_parameter_last_year(cpi_u)
 
 
 def extend_single_year_dataset(
     dataset: USSingleYearDataset,
-    end_year: int = DEFAULT_END_YEAR,
+    end_year: int | None = None,
     system=None,
 ) -> USMultiYearDataset:
     """Extend a single-year US dataset to multiple years via uprating.
@@ -17,8 +47,19 @@ def extend_single_year_dataset(
     ``end_year``, then applies multiplicative uprating using growth factors
     derived from the policyengine-us parameter tree.
 
+    If ``end_year`` is not provided, it defaults to the latest year
+    covered by the CPI-U parameter (gov.bls.cpi.cpi_u).
+
     Variables without an uprating parameter are carried forward unchanged.
     """
+    if system is None:
+        from policyengine_us.system import system as _system
+
+        system = _system
+
+    if end_year is None:
+        end_year = _get_default_end_year(system)
+
     start_year = int(dataset.time_period)
     if end_year < start_year:
         raise ValueError(

…ixtures/economic_assumptions_fixtures.py …tures/test_extend_single_year_dataset.pypolicyengine_us/tests/microsimulation/data/fixtures/economic_assumptions_fixtures.py renamed to policyengine_us/tests/microsimulation/data/fixtures/test_extend_single_year_dataset.py policyengine_us/tests/microsimulation/data/fixtures/economic_assumptions_fixtures.py renamed to policyengine_us/tests/microsimulation/data/fixtures/test_extend_single_year_dataset.py

@@ -1,13 +1,14 @@
 """
-Fixtures for extend_single_year_dataset and uprating tests.
+Fixtures for test_extend_single_year_dataset.py.
 
-Provides mock system objects, parameter trees, and sample datasets
-so tests can run without loading the full policyengine-us tax-benefit
-system.
+Provides constants, mock system objects, parameter trees, sample datasets,
+and helper functions so tests can run without loading the full policyengine-us
+tax-benefit system.
 """
 
 import numpy as np
 import pandas as pd
+import pytest
 
 from policyengine_us.data.dataset_schema import USSingleYearDataset
 
@@ -17,6 +18,9 @@
 
 BASE_YEAR = 2024
 END_YEAR_DEFAULT = 2035
+# END_YEAR_SHORT caps how far tests extend datasets.  The mock parameter
+# fixtures only define values through 2026-2027, so this is the furthest
+# the mock system can uprate to without hitting a missing key.
 END_YEAR_SHORT = 2026
 
 NUM_PERSONS = 10
@@ -221,3 +225,63 @@ def build_single_year_dataset(
         marital_unit=pd.DataFrame({"marital_unit_id": [1, 2, 3, 4]}),
         time_period=time_period,
     )
+
+
+# ---------------------------------------------------------------------------
+# Helper to call extend_single_year_dataset with a mock system
+# ---------------------------------------------------------------------------
+
+
+def call_extend_with_mock_system(mock_system, dataset, **kwargs):
+    """Call extend_single_year_dataset passing mock system directly."""
+    from policyengine_us.data.economic_assumptions import (
+        extend_single_year_dataset,
+    )
+
+    return extend_single_year_dataset(dataset, system=mock_system, **kwargs)
+
+
+# ---------------------------------------------------------------------------
+# Mock Microsimulation.__init__ helpers
+# ---------------------------------------------------------------------------
+
+
+class MockHolder:
+    """Minimal holder stub — reports no known periods."""
+
+    def get_known_periods(self):
+        return []
+
+
+def make_mock_super_init(system_module, captured=None):
+    """Return a mock CoreMicrosimulation.__init__ that sets up just enough
+    state for the rest of Microsimulation.__init__ to run."""
+
+    def mock_super_init(self, *args, **kwargs):
+        ds = kwargs.get("dataset")
+        if captured is not None:
+            captured[0] = ds
+        self.dataset = ds
+        self.tax_benefit_system = system_module.system
+        self.is_over_dataset = True
+        self.input_variables = []
+        self.get_holder = lambda name: MockHolder()
+        self.set_input = lambda *a, **kw: None
+        self.apply_reform = lambda r: None
+
+    return mock_super_init
+
+
+# ---------------------------------------------------------------------------
+# Pytest fixtures
+# ---------------------------------------------------------------------------
+
+
+@pytest.fixture
+def mock_system():
+    return build_mock_system()
+
+
+@pytest.fixture
+def base_dataset():
+    return build_single_year_dataset()