Merge pull request #594 from bashtage/numpy-3

bashtage · web-flow · commit 31e703399032 · 2024-04-16T15:18:25.000+01:00
COMPAT: Add monkey patch for formulaic
diff --git a/ci/azure_template_posix.yml b/ci/azure_template_posix.yml
@@ -152,9 +152,8 @@ jobs:
       testRunTitle: 'Python $(python.version)'
     condition: succeededOrFailed()
 
-  - task: PublishCodeCoverageResults@1
+  - task: PublishCodeCoverageResults@2
     inputs:
-      codeCoverageTool: Cobertura
       summaryFileLocation: '$(System.DefaultWorkingDirectory)/**/coverage.xml'
     condition: and(eq(variables['coverage'], 'true'), ne(variables['test.install'], 'true'))
 
diff --git a/doc/source/conf.py b/doc/source/conf.py
@@ -10,7 +10,6 @@
 import glob
 import hashlib
 import os
-from typing import Dict, List
 
 from packaging.version import parse
 
diff --git a/examples/system_formulas.ipynb b/examples/system_formulas.ipynb
@@ -46,9 +46,9 @@
     "from collections import OrderedDict\n",
     "\n",
     "formula = OrderedDict()\n",
-    "formula[\n",
-    "    \"benefits\"\n",
-    "] = \"hrbens ~ educ + exper + expersq + union + south + nrtheast + nrthcen + male\"\n",
+    "formula[\"benefits\"] = (\n",
+    "    \"hrbens ~ educ + exper + expersq + union + south + nrtheast + nrthcen + male\"\n",
+    ")\n",
     "formula[\"earnings\"] = \"hrearn ~ educ + exper + expersq + nrtheast + married + male\""
    ]
   },
diff --git a/linearmodels/asset_pricing/covariance.py b/linearmodels/asset_pricing/covariance.py
@@ -1,6 +1,7 @@
 """
 Covariance estimators for linear factor models
 """
+
 from __future__ import annotations
 
 from numpy import empty, ndarray
diff --git a/linearmodels/asset_pricing/model.py b/linearmodels/asset_pricing/model.py
@@ -1,6 +1,7 @@
 """
 Linear factor models for applications in asset pricing
 """
+
 from __future__ import annotations
 
 from typing import Any, Callable, cast
diff --git a/linearmodels/asset_pricing/results.py b/linearmodels/asset_pricing/results.py
@@ -1,6 +1,7 @@
 """
 Results for linear factor models
 """
+
 from __future__ import annotations
 
 from linearmodels.compat.statsmodels import Summary
diff --git a/linearmodels/compat/formulaic.py b/linearmodels/compat/formulaic.py
@@ -0,0 +1,12 @@
+def monkey_patch_materializers():
+    from formulaic.materializers.base import FormulaMaterializer
+    from formulaic.materializers.pandas import PandasMaterializer
+
+    if "pandas.DataFrame" not in FormulaMaterializer.REGISTERED_INPUTS:
+        FormulaMaterializer.REGISTERED_INPUTS["pandas.DataFrame"] = (
+            FormulaMaterializer.REGISTERED_INPUTS["pandas.core.frame.DataFrame"]
+        )
+    if "pandas.DataFrame" not in PandasMaterializer.REGISTERED_INPUTS:
+        PandasMaterializer.REGISTERED_INPUTS["pandas.DataFrame"] = (
+            PandasMaterializer.REGISTERED_INPUTS["pandas.core.frame.DataFrame"]
+        )
diff --git a/linearmodels/iv/_utility.py b/linearmodels/iv/_utility.py
@@ -12,6 +12,11 @@
 
 from linearmodels.typing import Float64Array
 
+from ..compat.formulaic import monkey_patch_materializers
+
+# Monkey patch parsers if needed, remove once formulaic updated
+monkey_patch_materializers()
+
 PARSING_ERROR = """
 Conversion of formula blocks to DataFrames failed.
 The formula blocks used for conversion were:
diff --git a/linearmodels/iv/absorbing.py b/linearmodels/iv/absorbing.py
@@ -125,9 +125,9 @@ def lsmr_annihilate(
         return empty_like(y)
     use_cache = use_cache and x_hash is not None
     regressor_hash = x_hash if x_hash is not None else ""
-    default_opts: dict[
-        str, bool | float | str | ArrayLike | None | dict[str, Any]
-    ] = dict(atol=1e-8, btol=1e-8, show=False)
+    default_opts: dict[str, bool | float | str | ArrayLike | None | dict[str, Any]] = (
+        dict(atol=1e-8, btol=1e-8, show=False)
+    )
     assert lsmr_options is not None
     default_opts.update(lsmr_options)
     resids = []
@@ -835,8 +835,9 @@ def _prepare_interactions(self) -> None:
     def _first_time_fit(
         self,
         use_cache: bool,
-        absorb_options: None
-        | (dict[str, bool | float | str | ArrayLike | None | dict[str, Any]]),
+        absorb_options: None | (
+            dict[str, bool | float | str | ArrayLike | None | dict[str, Any]]
+        ),
         method: str,
     ) -> None:
         weights = (
@@ -947,8 +948,9 @@ def fit(
         cov_type: str = "robust",
         debiased: bool = False,
         method: str = "auto",
-        absorb_options: None
-        | (dict[str, bool | float | str | ArrayLike | None | dict[str, Any]]) = None,
+        absorb_options: None | (
+            dict[str, bool | float | str | ArrayLike | None | dict[str, Any]]
+        ) = None,
         use_cache: bool = True,
         lsmr_options: dict[str, float | bool] | None = None,
         **cov_config: Any,
diff --git a/linearmodels/iv/model.py b/linearmodels/iv/model.py
@@ -1,6 +1,7 @@
 """
 Instrumental variable estimators
 """
+
 from __future__ import annotations
 
 from typing import Any, TypeVar, Union, cast
diff --git a/linearmodels/iv/results.py b/linearmodels/iv/results.py
@@ -1,6 +1,7 @@
 """
 Results containers and post-estimation diagnostics for IV models
 """
+
 from __future__ import annotations
 
 from linearmodels.compat.statsmodels import Summary
diff --git a/linearmodels/panel/covariance.py b/linearmodels/panel/covariance.py
@@ -79,6 +79,7 @@ class HomoskedasticCovariance:
     where df is ``extra_df`` and n-df is replace by n-df-k if ``debiased`` is
     ``True``.
     """
+
     ALLOWED_KWARGS: tuple[str, ...] = tuple()
     DEFAULT_KERNEL = "newey-west"
 
@@ -270,6 +271,7 @@ class ClusteredCovariance(HomoskedasticCovariance):
     where g is the number of distinct groups and n is the number of
     observations.
     """
+
     ALLOWED_KWARGS = ("clusters", "group_debias")
 
     def __init__(
@@ -400,6 +402,7 @@ class DriscollKraay(HomoskedasticCovariance):
     where df is ``extra_df`` and n-df is replace by n-df-k if ``debiased`` is
     ``True``. :math:`K(i, bw)` is the kernel weighting function.
     """
+
     ALLOWED_KWARGS = ("kernel", "bandwidth")
     # TODO: Test
 
@@ -517,6 +520,7 @@ class ACCovariance(HomoskedasticCovariance):
     where df is ``extra_df`` and n-df is replace by n-df-k if ``debiased`` is
     ``True``. :math:`K(i, bw)` is the kernel weighting function.
     """
+
     ALLOWED_KWARGS = ("kernel", "bandwidth")
     # TODO: Docstring
 
diff --git a/linearmodels/panel/data.py b/linearmodels/panel/data.py
@@ -254,7 +254,7 @@ def __init__(
             raise TypeError("Only ndarrays, DataFrames or DataArrays are " "supported")
         if convert_dummies:
             self._frame = expand_categoricals(self._frame, drop_first)
-            self._frame = self._frame.astype(np.float64, copy=False)
+            self._frame = self._frame.astype(np.float64)
 
         time_index = Series(self.index.levels[1])
         if not (
@@ -519,32 +519,29 @@ def demean_pass(
         return PanelData(current)
 
     @overload
-    def demean(
+    def demean(  # noqa: E704
         self,
         group: Literal["entity", "time", "both"],
         *,
         return_panel: Literal[False],
-    ) -> Float64Array:
-        ...
+    ) -> Float64Array: ...
 
     @overload
-    def demean(
+    def demean(  # noqa: E704
         self,
         group: Literal["entity", "time", "both"] = ...,
         weights: PanelData | None = ...,
         return_panel: Literal[True] = ...,
         low_memory: bool = ...,
-    ) -> PanelData:
-        ...
+    ) -> PanelData: ...
 
     @overload
-    def demean(
+    def demean(  # noqa: E704
         self,
         group: Literal["entity", "time", "both"],
         weights: PanelData | None,
         return_panel: Literal[False],
-    ) -> Float64Array:
-        ...
+    ) -> Float64Array: ...  # noqa: E704
 
     def demean(
         self,
@@ -757,7 +754,7 @@ def dummies(self, group: str = "entity", drop_first: bool = False) -> DataFrame:
         cols = self.entities if group == "entity" else self.time
         # TODO: Incorrect typing in pandas-stubs not handling Hashable | None
         dummy_cols = [c for c in cols if c in dummies]
-        return dummies[dummy_cols].astype(np.float64, copy=False)  # type: ignore
+        return dummies[dummy_cols].astype(np.float64)  # type: ignore
 
 
 PanelDataLike = Union[PanelData, ArrayLike]
diff --git a/linearmodels/panel/model.py b/linearmodels/panel/model.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from linearmodels.compat.formulaic import monkey_patch_materializers
+
 from collections.abc import Mapping
 from typing import Any, NamedTuple, Union, cast
 
@@ -61,6 +63,9 @@
     NumericArray,
 )
 
+# Monkey patch parsers if needed, remove once formulaic updated
+monkey_patch_materializers()
+
 CovarianceEstimator = Union[
     ACCovariance,
     ClusteredCovariance,
diff --git a/linearmodels/shared/hypotheses.py b/linearmodels/shared/hypotheses.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from linearmodels.compat.formulaic import monkey_patch_materializers
+
 from collections.abc import Mapping
 
 from formulaic.utils.constraints import LinearConstraints
@@ -9,6 +11,9 @@
 
 from linearmodels.typing import ArrayLike
 
+# Monkey patch parsers if needed, remove once formulaic updated
+monkey_patch_materializers()
+
 
 class WaldTestStatistic:
     """
diff --git a/linearmodels/shared/utility.py b/linearmodels/shared/utility.py
@@ -23,11 +23,9 @@
 
 
 class SupportsKeysAndGetItem(Protocol[_KT, _VT_co]):
-    def keys(self) -> Iterable[_KT]:
-        ...
+    def keys(self) -> Iterable[_KT]: ...  # noqa: E704
 
-    def __getitem__(self, __k: _KT) -> _VT_co:
-        ...
+    def __getitem__(self, __k: _KT) -> _VT_co: ...  # noqa: E704
 
 
 def _new_attr_dict_(*args: Iterable[tuple[Any, Any]]) -> AttrDict:
diff --git a/linearmodels/system/gmm.py b/linearmodels/system/gmm.py
@@ -1,6 +1,7 @@
 """
 Covariance and weight estimation for GMM IV estimators
 """
+
 from __future__ import annotations
 
 from collections.abc import Sequence
diff --git a/linearmodels/system/model.py b/linearmodels/system/model.py
@@ -11,8 +11,11 @@
     Systems of Simultaneous Equations in R. Journal of Statistical Software,
     23(4), 1 - 40. doi:http://dx.doi.org/10.18637/jss.v023.i04
 """
+
 from __future__ import annotations
 
+from linearmodels.compat.formulaic import monkey_patch_materializers
+
 from collections.abc import Mapping, Sequence
 from functools import reduce
 import textwrap
@@ -57,6 +60,9 @@
 
 __all__ = ["SUR", "IV3SLS", "IVSystemGMM", "LinearConstraint"]
 
+# Monkey patch parsers if needed, remove once formulaic updated
+monkey_patch_materializers()
+
 UNKNOWN_EQ_TYPE = """
 Contents of each equation must be either a dictionary with keys "dependent"
 and "exog" or a 2-element tuple of he form (dependent, exog).
@@ -877,8 +883,7 @@ def _common_indiv_results(
         constant: bool,
         total_ss: float,
         *,
-        weight_est: None
-        | (
+        weight_est: None | (
             HomoskedasticWeightMatrix | HeteroskedasticWeightMatrix | KernelWeightMatrix
         ) = None,
     ) -> AttrDict:
diff --git a/linearmodels/tests/iv/results/simulated-test-data.py b/linearmodels/tests/iv/results/simulated-test-data.py
@@ -12,6 +12,7 @@
 Will also test other configurations - small, covariance-estimators, constant
 
 """
+
 import numpy as np
 from numpy.random import multivariate_normal, seed
 import pandas as pd
diff --git a/linearmodels/tests/system/results/execute-stata-3sls.py b/linearmodels/tests/system/results/execute-stata-3sls.py
@@ -2,6 +2,7 @@
 Important cases
 
 """
+
 import os
 import subprocess
 
diff --git a/linearmodels/tests/system/results/execute-stata.py b/linearmodels/tests/system/results/execute-stata.py
@@ -5,6 +5,7 @@
 2. Small sample adjustment
 3. Constraints across equations
 """
+
 import os
 import subprocess
 
diff --git a/linearmodels/typing/__init__.py b/linearmodels/typing/__init__.py
@@ -48,6 +48,6 @@
         np.ndarray[Any, np.dtype[np.floating[Any]]],  # pragma: no cover
     ]  # pragma: no cover
 else:
-    IntArray = (
-        Float64Array
-    ) = Int64Array = Int32Array = BoolArray = AnyArray = NumericArray = np.ndarray
+    IntArray = Float64Array = Int64Array = Int32Array = BoolArray = AnyArray = (
+        NumericArray
+    ) = np.ndarray
diff --git a/linearmodels/typing/data.py b/linearmodels/typing/data.py
@@ -29,9 +29,9 @@
     AnyArray = np.ndarray[Any, Any]  # pragma: no cover
     Uint32Array = np.ndarray[Any, np.dtype[np.uint32]]  # pragma: no cover
 else:
-    Uint32Array = (
-        IntArray
-    ) = Float64Array = Int64Array = Int32Array = BoolArray = AnyArray = NDArray
+    Uint32Array = IntArray = Float64Array = Int64Array = Int32Array = BoolArray = (
+        AnyArray
+    ) = NDArray
 
 __all__ = [
     "Float64Array",
diff --git a/setup.py b/setup.py
@@ -1,10 +1,9 @@
 from setuptools import Extension, find_namespace_packages, setup
 from setuptools.dist import Distribution
+from setuptools.errors import CCompilerError, ExecError, PlatformError
 
-from distutils.errors import CCompilerError, DistutilsExecError, DistutilsPlatformError
 import glob
 import os
-from typing import Dict
 
 try:
     from Cython.Build import cythonize
@@ -136,8 +135,8 @@ def run_setup(binary: bool = True) -> None:
     run_setup(binary=build_binary)
 except (
     CCompilerError,
-    DistutilsExecError,
-    DistutilsPlatformError,
+    ExecError,
+    PlatformError,
     OSError,
     ValueError,
     ImportError,
