TYP: Improve typing

doc/source/conf.py

@@ -90,7 +90,7 @@
     full_name = os.path.split(file_to_copy)[-1]
     folder, file_name = full_name.split("_")
     if not file_name.endswith("ipynb"):
-        file_name = "_".join((folder, file_name))
+        file_name = f"{folder}_{file_name}"
     out_dir = os.path.join(folder, "examples")
     if not os.path.exists(out_dir):
         os.makedirs(out_dir, exist_ok=True)

linearmodels/__init__.py

@@ -56,10 +56,10 @@
 from .system import IV3SLS, SUR, IVSystemGMM
 
 OLS = _OLS
-WARN_ON_MISSING = os.environ.get("LINEARMODELS_WARN_ON_MISSING", "1")
-WARN_ON_MISSING = False if WARN_ON_MISSING in ("", "0", "false", "False") else True
-DROP_MISSING = os.environ.get("LINEARMODELS_DROP_MISSING", "1")
-DROP_MISSING = False if DROP_MISSING in ("", "0", "false", "False") else True
+_WARN_ON_MISSING = os.environ.get("LINEARMODELS_WARN_ON_MISSING", "1")
+WARN_ON_MISSING = False if _WARN_ON_MISSING in ("", "0", "false", "False") else True
+_DROP_MISSING = os.environ.get("LINEARMODELS_DROP_MISSING", "1")
+DROP_MISSING = False if _DROP_MISSING in ("", "0", "false", "False") else True
 
 __all__ = [
     "DROP_MISSING",

linearmodels/conftest.py

@@ -20,7 +20,7 @@
     logger.critical("Copy on Write testing enabled")
 
 
-def pytest_configure(config):
+def pytest_configure(config: pytest.Config) -> None:
     # Minimal config to simplify running tests from lm.test()
     config.addinivalue_line("markers", "example: mark a test as an example")
     config.addinivalue_line("markers", "slow: mark a test as slow")
@@ -32,15 +32,15 @@ def pytest_configure(config):
     )
 
 
-def pytest_addoption(parser):
+def pytest_addoption(parser: pytest.Parser) -> None:
     parser.addoption("--skip-slow", action="store_true", help="skip slow tests")
     parser.addoption("--only-slow", action="store_true", help="run only slow tests")
     parser.addoption("--skip-smoke", action="store_true", help="skip smoke tests")
     parser.addoption("--only-smoke", action="store_true", help="run only smoke tests")
     parser.addoption("--skip-examples", action="store_true", help="skip examples tests")
 
 
-def pytest_runtest_setup(item):
+def pytest_runtest_setup(item: pytest.Item) -> None:
     if "slow" in item.keywords and item.config.getoption("--skip-slow"):
         pytest.skip("skipping due to --skip-slow")
 

linearmodels/iv/absorbing.py

@@ -77,7 +77,7 @@ def __init__(self) -> None:
             self._hasher = hashlib.sha256()
             self._use_xxh64 = False
 
-    def reset(self):
+    def reset(self) -> None:
         if self._use_xxh64:
             assert isinstance(self._hasher, xxh64)
             self._hasher.reset()
@@ -108,7 +108,7 @@ def lsmr_annihilate(
     use_cache: bool = True,
     x_hash: Hashable | None = None,
     **lsmr_options: (
-        bool | float | str | linearmodels.typing.data.ArrayLike | None | dict[str, Any]
+        bool | float | str | linearmodels.typing.data.ArrayLike | dict[str, Any] | None
     ),
 ) -> linearmodels.typing.data.Float64Array:
     r"""
@@ -149,7 +149,7 @@ def lsmr_annihilate(
     regressor_hash = x_hash if x_hash is not None else ""
     default_opts: dict[
         str,
-        bool | float | str | linearmodels.typing.data.ArrayLike | None | dict[str, Any],
+        bool | float | str | linearmodels.typing.data.ArrayLike | dict[str, Any] | None,
     ] = {"atol": 1e-8, "btol": 1e-8, "show": False}
     assert lsmr_options is not None
     default_opts.update(lsmr_options)
@@ -865,16 +865,17 @@ def _prepare_interactions(self) -> None:
     def _first_time_fit(
         self,
         use_cache: bool,
-        absorb_options: None | (
+        absorb_options: (
             dict[
                 str,
                 bool
                 | float
                 | str
                 | linearmodels.typing.data.ArrayLike
-                | None
-                | dict[str, Any],
+                | dict[str, Any]
+                | None,
             ]
+            | None
         ),
         method: str,
     ) -> None:
@@ -988,16 +989,17 @@ def fit(
         cov_type: str = "robust",
         debiased: bool = False,
         method: str = "auto",
-        absorb_options: None | (
+        absorb_options: (
             dict[
                 str,
                 bool
                 | float
                 | str
                 | linearmodels.typing.data.ArrayLike
-                | None
-                | dict[str, Any],
+                | dict[str, Any]
+                | None,
             ]
+            | None
         ) = None,
         use_cache: bool = True,
         lsmr_options: dict[str, float | bool] | None = None,

linearmodels/iv/results.py

@@ -584,7 +584,7 @@ def predict(
         return out_df
 
     @property
-    def kappa(self) -> float:
+    def kappa(self) -> float | None:
         """k-class estimator value"""
         return self._kappa
 

linearmodels/panel/data.py

@@ -358,7 +358,7 @@ def entities(self) -> list[linearmodels.typing.Label]:
         return list(index.levels[0][index.codes[0]].unique())
 
     @property
-    def entity_ids(self) -> linearmodels.typing.data.IntArray:
+    def entity_ids(self) -> linearmodels.typing.data.AnyIntArray:
         """
         Get array containing entity group membership information
 
@@ -371,7 +371,7 @@ def entity_ids(self) -> linearmodels.typing.data.IntArray:
         return np.asarray(index.codes[0])[:, None]
 
     @property
-    def time_ids(self) -> linearmodels.typing.data.IntArray:
+    def time_ids(self) -> linearmodels.typing.data.AnyIntArray:
         """
         Get array containing time membership information
 

linearmodels/shared/hypotheses.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 from collections.abc import Mapping
+from typing import Any
 
 from formulaic.utils.constraints import LinearConstraints
 import numpy as np
@@ -46,6 +47,7 @@ def __init__(
         self.df = df
         self.df_denom = df_denom
         self._name = name
+        self.dist: Any
         if df_denom is None:
             self.dist = chi2(df)
             self.dist_name = f"chi2({df})"

linearmodels/system/model.py

@@ -909,8 +909,11 @@ def _common_indiv_results(
         constant: bool,
         total_ss: float,
         *,
-        weight_est: None | (
-            HomoskedasticWeightMatrix | HeteroskedasticWeightMatrix | KernelWeightMatrix
+        weight_est: (
+            HomoskedasticWeightMatrix
+            | HeteroskedasticWeightMatrix
+            | KernelWeightMatrix
+            | None
         ) = None,
     ) -> AttrDict:
         loc = 0

linearmodels/tests/asset_pricing/test_linear_factor_gmm.py

@@ -78,19 +78,21 @@ def test_linear_model_gmm_smoke_risk_free(data):
     mod = LinearFactorModelGMM(data.portfolios, data.factors, risk_free=True)
     res = mod.fit(cov_type="robust", disp=10)
     get_all(res)
-    str(res._cov_est)
-    res._cov_est.__repr__()
-    str(res._cov_est.config)
+    # Smoke tests
+    assert isinstance(str(res._cov_est), str)
+    assert isinstance(res._cov_est.__repr__(), str)
+    assert isinstance(str(res._cov_est.config), str)
 
 
 @pytest.mark.smoke
 def test_linear_model_gmm_kernel_smoke(data):
     mod = LinearFactorModelGMM(data.portfolios, data.factors)
     res = mod.fit(cov_type="kernel", disp=10)
     get_all(res)
-    str(res._cov_est)
-    res._cov_est.__repr__()
-    str(res._cov_est.config)
+    # Smoke tests
+    assert isinstance(str(res._cov_est), str)
+    assert isinstance(res._cov_est.__repr__(), str)
+    assert isinstance(str(res._cov_est.config), str)
 
 
 @pytest.mark.smoke

linearmodels/tests/iv/test_model.py

@@ -394,10 +394,10 @@ def test_first_stage_summary(data):
 
 def test_gmm_str(data):
     mod = IVGMM(data.dep, data.exog, data.endog, data.instr)
-    str(mod.fit(cov_type="unadjusted"))
-    str(mod.fit(cov_type="robust"))
-    str(mod.fit(cov_type="clustered", clusters=data.clusters))
-    str(mod.fit(cov_type="kernel"))
+    assert isinstance(str(mod.fit(cov_type="unadjusted")), str)
+    assert isinstance(str(mod.fit(cov_type="robust")), str)
+    assert isinstance(str(mod.fit(cov_type="clustered", clusters=data.clusters)), str)
+    assert isinstance(str(mod.fit(cov_type="kernel")), str)
 
 
 def test_gmm_cue_optimization_options(small_data):

linearmodels/tests/iv/test_results.py

@@ -34,7 +34,7 @@ def result_checker(res):
             _attr()
         else:
             assert isinstance(_attr, object)
-            str(_attr)
+            assert isinstance(str(_attr), str)
 
 
 def test_results(data, model):

linearmodels/tests/system/results/execute-stata-3sls.py

@@ -45,10 +45,8 @@
 STATA_PATH = os.path.join("C:\\", "Program Files (x86)", "Stata15", "StataMP-64.exe")
 OUTFILE = os.path.join(os.getcwd(), "stata-3sls-results.txt")
 
-header = [
-    r'use "C:\git\linearmodels\linearmodels\tests\system\results\simulated-3sls.dta",'
-    r" clear"
-]
+header = r'use "C:\git\linearmodels\linearmodels\tests\system\results\simulated-3sls.dta", clear'
+
 
 all_stats = (
     "estout using {outfile}, cells(b(fmt(%13.12g)) t(fmt(%13.12g)) "

linearmodels/tests/system/results/execute-stata.py

@@ -16,10 +16,7 @@
 STATA_PATH = os.path.join("C:\\", "Program Files (x86)", "Stata13", "StataMP-64.exe")
 OUTFILE = os.path.join(os.getcwd(), "stata-sur-results.txt")
 
-header = [
-    r'use "C:\git\linearmodels\linearmodels\tests\system\results\simulated-sur.dta"'
-    ", clear"
-]
+header = r'use "C:\git\linearmodels\linearmodels\tests\system\results\simulated-sur.dta", clear'
 
 all_stats = (
     "estout using {outfile}, cells(b(fmt(%13.12g)) "

linearmodels/typing/data.py

@@ -4,12 +4,14 @@
 import pandas as pd
 
 base_data_types = [np.ndarray, pd.DataFrame, pd.Series]
+
 try:
     import xarray as xr
 
     ArrayLike = Union[np.ndarray, xr.DataArray, pd.DataFrame, pd.Series]
 
 except ImportError:
+
     # Always needed to allow optional xarray
     ArrayLike = Union[np.ndarray, pd.DataFrame, pd.Series]
 
@@ -20,13 +22,15 @@
 Int64Array = np.ndarray[tuple[int, ...], np.dtype[np.int64]]  # pragma: no cover
 Int32Array = np.ndarray[tuple[int, ...], np.dtype[np.int32]]  # pragma: no cover
 IntArray = np.ndarray[tuple[int, ...], np.dtype[np.int_]]  # pragma: no cover
+AnyIntArray = np.ndarray[tuple[int, ...], np.dtype[np.integer]]  # pragma: no cover
 BoolArray = np.ndarray[tuple[int, ...], np.dtype[np.bool_]]  # pragma: no cover
 AnyArray = np.ndarray[tuple[int, ...], Any]  # pragma: no cover
 Uint32Array = np.ndarray[tuple[int, ...], np.dtype[np.uint32]]  # pragma: no cover
 FloatArray1D = np.ndarray[tuple[int], np.dtype[np.float64]]
 FloatArray2D = np.ndarray[tuple[int, int], np.dtype[np.float64]]
 __all__ = [
     "AnyArray",
+    "AnyIntArray",
     "ArrayLike",
     "BoolArray",
     "Float64Array",

pyproject.toml

@@ -126,13 +126,13 @@ dev = [
   "pytest-xdist",
   "pytest-cov",
   # formatting
-  "black[jupyter]~=25.9.0",
-  "isort~=6.0",
+  "black[jupyter]~=26.5.0",
+  "isort~=9.0",
   "colorama",
   "flake8",
   "flake8-bugbear",
   "mypy>=1.3",
-  "ruff>=0.8.6",
+  "ruff>=0.16.0",
   "pyupgrade>=3.4.0",
   "jupyterlab-code-formatter",
   "jupyterlab>=4.4.8",  # not directly required, pinned by Snyk to avoid a vulnerability
@@ -153,7 +153,7 @@ setup = ['--vsenv']
 version_file = "arch/_version.py"
 
 [tool.black]
-target-version = ['py310', 'py311', 'py312', 'py313']
+target-version = ['py310', 'py311', 'py312', 'py313', 'py314']
 exclude = '''
 (
   \.egg
@@ -256,6 +256,8 @@ ignore = [
   "PLR0915", # 21
   # Magic number
   "PLR2004",
+  # Too many positional arguments
+  "PLR0917",
   # Like to suggest use [a, *b] instead of [a] + b
   "RUF005",
 ]
@@ -376,7 +378,6 @@ omit = [
 directory = "coverage_html_report"
 
 [tool.mypy]
-plugins="numpy.typing.mypy_plugin"
 exclude = [
   "tests",
 ]

requirements-dev.txt

@@ -20,9 +20,9 @@ matplotlib
 
 # Linting
 mypy>=1.3
-black[jupyter]~=25.9.0
-isort>=5.12
-ruff>=0.8.6
+black[jupyter]~=26.5.0
+isort>=9.0
+ruff>=0.16.0
 flake8
 flake8-bugbear
 pandas-stubs

requirements-test.txt

@@ -1,10 +1,11 @@
-black[jupyter]~=25.9.0
+black[jupyter]~=26.5.0
 coverage
 flake8
 isort
+ruff
 colorama
 matplotlib
-pytest>=8.4.1,<9
+pytest>=9,<10
 pytest-xdist
 pytest-cov
 pytest-randomly