src/quant_engine/alpha_factors.py

@@ -3271,6 +3271,143 @@ def evaluate_phase3_formula(name: str, **inputs: pd.Series) -> pd.Series:
     return function(*(inputs[field] for field in required_inputs))
 
 
+# ── Phase 4 formula contract: frozen alpha051-alpha100 surface ──────────────
+
+# Phase 4 extends the versioned formula contract without mutating the Phase 3
+# catalogue, digest, dispatch surface, or the existing formula functions.
+ALPHA158_PHASE4_FORMULA_CONTRACT_VERSION = "1.0.0"
+ALPHA158_PHASE4_FORMULA_CATALOG_SHA256 = (
+    "858daf5e2abab5063fc28fcf7c79936096e2c76dde582fc7bab78b3458f17054"
+)
+
+_PHASE4_FORMULA_FUNCTIONS: dict[str, Callable[..., pd.Series]] = {
+    "alpha_051": alpha_051,
+    "alpha_052": alpha_052,
+    "alpha_053": alpha_053,
+    "alpha_054": alpha_054,
+    "alpha_055": alpha_055,
+    "alpha_056": alpha_056,
+    "alpha_057": alpha_057,
+    "alpha_058": alpha_058,
+    "alpha_059": alpha_059,
+    "alpha_060": alpha_060,
+    "alpha_061": alpha_061,
+    "alpha_062": alpha_062,
+    "alpha_063": alpha_063,
+    "alpha_064": alpha_064,
+    "alpha_065": alpha_065,
+    "alpha_066": alpha_066,
+    "alpha_067": alpha_067,
+    "alpha_068": alpha_068,
+    "alpha_069": alpha_069,
+    "alpha_070": alpha_070,
+    "alpha_071": alpha_071,
+    "alpha_072": alpha_072,
+    "alpha_073": alpha_073,
+    "alpha_074": alpha_074,
+    "alpha_075": alpha_075,
+    "alpha_076": alpha_076,
+    "alpha_077": alpha_077,
+    "alpha_078": alpha_078,
+    "alpha_079": alpha_079,
+    "alpha_080": alpha_080,
+    "alpha_081": alpha_081,
+    "alpha_082": alpha_082,
+    "alpha_083": alpha_083,
+    "alpha_084": alpha_084,
+    "alpha_085": alpha_085,
+    "alpha_086": alpha_086,
+    "alpha_087": alpha_087,
+    "alpha_088": alpha_088,
+    "alpha_089": alpha_089,
+    "alpha_090": alpha_090,
+    "alpha_091": alpha_091,
+    "alpha_092": alpha_092,
+    "alpha_093": alpha_093,
+    "alpha_094": alpha_094,
+    "alpha_095": alpha_095,
+    "alpha_096": alpha_096,
+    "alpha_097": alpha_097,
+    "alpha_098": alpha_098,
+    "alpha_099": alpha_099,
+    "alpha_100": alpha_100,
+}
+
+_PHASE4_INPUT_CATEGORIES = {
+    1: "single",
+    2: "pair",
+    3: "triple",
+    4: "quadruple",
+    5: "quintuple",
+}
+
+
+def _build_phase4_formula_specs() -> dict[str, dict[str, Any]]:
+    specs: dict[str, dict[str, Any]] = {}
+    for alpha_id, function in _PHASE4_FORMULA_FUNCTIONS.items():
+        meta = ALPHA158_REGISTRY[alpha_id]
+        call_inputs = _phase3_call_inputs(function)
+        formula_inputs = _phase3_string_list(meta, "inputs", alpha_id)
+        input_category = _PHASE4_INPUT_CATEGORIES.get(len(call_inputs))
+        if input_category is None:
+            raise RuntimeError(f"unsupported formula input count for {alpha_id}")
+        specs[alpha_id] = {
+            "name": alpha_id,
+            "contract_version": ALPHA158_PHASE4_FORMULA_CONTRACT_VERSION,
+            "formula": meta["formula"],
+            "category": meta["category"],
+            "complexity": meta["complexity"],
+            "parameters": _phase3_string_list(meta, "params", alpha_id),
+            "description": meta["description"],
+            "references": _phase3_string_list(meta, "references", alpha_id),
+            "call_inputs": list(call_inputs),
+            "formula_inputs": formula_inputs,
+            "input_category": input_category,
+        }
+    return specs
+
+
+ALPHA158_PHASE4_FORMULA_SPECS: Mapping[str, Mapping[str, Any]] = (
+    _freeze_phase3_formula_specs(_build_phase4_formula_specs())
+)
+
+
+def list_phase4_formulas() -> tuple[str, ...]:
+    """Return the frozen alpha051-alpha100 formula IDs in stable order."""
+    return tuple(ALPHA158_PHASE4_FORMULA_SPECS)
+
+
+def evaluate_phase4_formula(name: str, **inputs: pd.Series) -> pd.Series:
+    """Evaluate a Phase 4 formula with an exact, alignment-safe input contract."""
+    if name not in ALPHA158_PHASE4_FORMULA_SPECS:
+        raise KeyError(f"formula {name!r} not registered")
+
+    spec = ALPHA158_PHASE4_FORMULA_SPECS[name]
+    required_inputs = cast(tuple[str, ...], spec["call_inputs"])
+    missing_inputs = [field for field in required_inputs if field not in inputs]
+    unexpected_inputs = sorted(field for field in inputs if field not in required_inputs)
+    if missing_inputs or unexpected_inputs:
+        details: list[str] = []
+        if missing_inputs:
+            details.append(f"missing inputs {missing_inputs}")
+        if unexpected_inputs:
+            details.append(f"unexpected inputs {unexpected_inputs}")
+        raise ValueError(f"invalid inputs for {name}: {'; '.join(details)}")
+
+    for field in required_inputs:
+        if not isinstance(inputs[field], pd.Series):
+            raise TypeError(f"{field} must be a pandas Series")
+
+    primary_field = required_inputs[0]
+    primary = inputs[primary_field]
+    for field in required_inputs[1:]:
+        if not primary.index.equals(inputs[field].index):
+            raise ValueError(f"{field} index must align with {primary_field}")
+
+    function = _PHASE4_FORMULA_FUNCTIONS[name]
+    return function(*(inputs[field] for field in required_inputs))
+
+
 __all__ = [
     "rank",
     "delta",
@@ -3309,6 +3446,11 @@ __all__ = [
     "ALPHA158_PHASE3_FORMULA_SPECS",
     "list_phase3_formulas",
     "evaluate_phase3_formula",
+    "ALPHA158_PHASE4_FORMULA_CONTRACT_VERSION",
+    "ALPHA158_PHASE4_FORMULA_CATALOG_SHA256",
+    "ALPHA158_PHASE4_FORMULA_SPECS",
+    "list_phase4_formulas",
+    "evaluate_phase4_formula",
     "alpha_001",
     "alpha_002",
     "alpha_003",

tests/test_alpha_factors.py

@@ -14,6 +14,9 @@ from quant_engine.alpha_factors import (
     ALPHA158_PHASE3_FORMULA_CATALOG_SHA256,
     ALPHA158_PHASE3_FORMULA_CONTRACT_VERSION,
     ALPHA158_PHASE3_FORMULA_SPECS,
+    ALPHA158_PHASE4_FORMULA_CATALOG_SHA256,
+    ALPHA158_PHASE4_FORMULA_CONTRACT_VERSION,
+    ALPHA158_PHASE4_FORMULA_SPECS,
     alpha_001,
     alpha_002,
     alpha_003,
@@ -175,9 +178,11 @@ from quant_engine.alpha_factors import (
     evaluate_phase1_operator,
     evaluate_phase2_operator,
     evaluate_phase3_formula,
+    evaluate_phase4_formula,
     list_phase1_operators,
     list_phase2_operators,
     list_phase3_formulas,
+    list_phase4_formulas,
     correlation,
     covariance,
     decay_linear,
@@ -1641,3 +1646,146 @@ def test_phase3_dispatch_rejects_implicit_series_alignment():
             close=inputs["close"],
             volume=misaligned_volume,
         )
+
+
+# ── Alpha158 Phase 4: versioned alpha051-alpha100 formula contract ──────────
+
+
+def test_phase4_formula_catalog_is_versioned_exact_and_content_addressed():
+    import hashlib
+    import json
+    from collections import Counter
+
+    expected_ids = tuple(f"alpha_{number:03d}" for number in range(51, 101))
+    expected_fields = {
+        "name",
+        "contract_version",
+        "formula",
+        "category",
+        "complexity",
+        "parameters",
+        "description",
+        "references",
+        "call_inputs",
+        "formula_inputs",
+        "input_category",
+    }
+
+    assert ALPHA158_PHASE4_FORMULA_CONTRACT_VERSION == "1.0.0"
+    assert list_phase4_formulas() == expected_ids
+    assert tuple(ALPHA158_PHASE4_FORMULA_SPECS) == expected_ids
+    assert Counter(
+        spec["input_category"] for spec in ALPHA158_PHASE4_FORMULA_SPECS.values()
+    ) == {"single": 1, "pair": 27, "triple": 11, "quadruple": 10, "quintuple": 1}
+
+    for alpha_id, spec in ALPHA158_PHASE4_FORMULA_SPECS.items():
+        assert set(spec) == expected_fields
+        assert spec["name"] == alpha_id
+        assert spec["contract_version"] == ALPHA158_PHASE4_FORMULA_CONTRACT_VERSION
+        assert spec["formula"] == ALPHA158_REGISTRY[alpha_id]["formula"]
+
+    serializable_specs = {
+        alpha_id: {
+            field: list(value) if isinstance(value, tuple) else value
+            for field, value in spec.items()
+        }
+        for alpha_id, spec in ALPHA158_PHASE4_FORMULA_SPECS.items()
+    }
+    encoded = json.dumps(
+        serializable_specs,
+        sort_keys=True,
+        separators=(",", ":"),
+        ensure_ascii=False,
+    ).encode()
+    assert hashlib.sha256(encoded).hexdigest() == ALPHA158_PHASE4_FORMULA_CATALOG_SHA256
+    assert ALPHA158_PHASE4_FORMULA_CATALOG_SHA256 == (
+        "858daf5e2abab5063fc28fcf7c79936096e2c76dde582fc7bab78b3458f17054"
+    )
+
+
+def test_phase4_formula_catalog_is_recursively_immutable():
+    import operator
+
+    with pytest.raises(TypeError):
+        operator.setitem(ALPHA158_PHASE4_FORMULA_SPECS, "alpha_051", {})
+    with pytest.raises(TypeError):
+        operator.setitem(
+            ALPHA158_PHASE4_FORMULA_SPECS["alpha_051"],
+            "formula",
+            "changed",
+        )
+    with pytest.raises(TypeError):
+        operator.setitem(
+            ALPHA158_PHASE4_FORMULA_SPECS["alpha_051"]["call_inputs"],
+            0,
+            "volume",
+        )
+
+
+def test_phase4_catalog_freezes_callable_and_formula_inputs():
+    import inspect
+
+    for alpha_id, spec in ALPHA158_PHASE4_FORMULA_SPECS.items():
+        function = getattr(alpha_factors_module, alpha_id)
+        signature_inputs = tuple(
+            "open" if name == "open_" else name
+            for name in inspect.signature(function).parameters
+        )
+        assert spec["call_inputs"] == signature_inputs
+        assert spec["formula_inputs"] == tuple(ALPHA158_REGISTRY[alpha_id]["inputs"])
+
+    assert ALPHA158_PHASE4_FORMULA_SPECS["alpha_055"]["call_inputs"] == (
+        "open",
+        "high",
+        "low",
+        "volume",
+        "close",
+    )
+
+
+def test_phase4_dispatch_matches_all_existing_alpha051_alpha100_functions():
+    inputs = _phase3_market_inputs()
+
+    for alpha_id, spec in ALPHA158_PHASE4_FORMULA_SPECS.items():
+        call_inputs = spec["call_inputs"]
+        function = getattr(alpha_factors_module, alpha_id)
+        expected = function(*(inputs[name] for name in call_inputs))
+        actual = evaluate_phase4_formula(
+            alpha_id,
+            **{name: inputs[name] for name in reversed(call_inputs)},
+        )
+        pd.testing.assert_series_equal(actual, expected)
+
+
+def test_phase4_dispatch_rejects_unknown_missing_extra_and_non_series_inputs():
+    inputs = _phase3_market_inputs()
+
+    with pytest.raises(KeyError, match="not registered"):
+        evaluate_phase4_formula("alpha_050", close=inputs["close"])
+    with pytest.raises(ValueError, match=r"missing inputs.*low"):
+        evaluate_phase4_formula("alpha_051", high=inputs["high"])
+    with pytest.raises(ValueError, match=r"unexpected inputs.*vwap"):
+        evaluate_phase4_formula(
+            "alpha_051",
+            high=inputs["high"],
+            low=inputs["low"],
+            vwap=inputs["vwap"],
+        )
+    with pytest.raises(TypeError, match="high must be a pandas Series"):
+        evaluate_phase4_formula(  # type: ignore[arg-type]
+            "alpha_051",
+            high=[1.0, 2.0],
+            low=inputs["low"],
+        )
+
+
+def test_phase4_dispatch_rejects_implicit_series_alignment():
+    inputs = _phase3_market_inputs()
+    misaligned_low = inputs["low"].rename(index={79: 80})
+
+    with pytest.raises(ValueError, match="low index must align with high"):
+        evaluate_phase4_formula(
+            "alpha_051",
+            high=inputs["high"],
+            low=misaligned_low,
+        )