Merge pull request #833 from bashtage/lm71-bug7-small-fixes

BUG: Reject an unknown group in PanelData.mean, count, demean and dum…

Author: bashtage

linearmodels/asset_pricing/covariance.py

@@ -4,7 +4,7 @@
 
 from __future__ import annotations
 
-from numpy import empty, ndarray, zeros_like
+from numpy import empty, errstate, ndarray, zeros_like
 from numpy.linalg import inv
 
 from linearmodels.iv.covariance import (
@@ -36,7 +36,10 @@ def bandwidth(self) -> float:
         if self._bandwidth is None:
             assert self._moments is not None
             moments = self._moments
-            m = moments / moments.std(0)[None, :]
+            # A moment that does not vary has a standard deviation of 0, and the
+            # series is NaN. kernel_optimal_bandwidth raises, so do not warn
+            with errstate(divide="ignore", invalid="ignore"):
+                m = moments / moments.std(0)[None, :]
             m = m.sum(1)
             bw = kernel_optimal_bandwidth(m, kernel=self.kernel)
             self._bandwidth = bw

linearmodels/asset_pricing/model.py

@@ -25,10 +25,16 @@
     GMMFactorModelResults,
     LinearFactorModelResults,
 )
+from linearmodels.iv.covariance import KERNEL_LOOKUP
 from linearmodels.iv.data import IVData, IVDataLike
 from linearmodels.shared import hypotheses
 from linearmodels.shared.alignment import check_alignment, row_labels
-from linearmodels.shared.exceptions import missing_warning, unknown_option_error
+from linearmodels.shared.exceptions import (
+    check_cov_config,
+    lower_option,
+    missing_warning,
+    unknown_option_error,
+)
 from linearmodels.shared.hypotheses import WaldTestStatistic
 from linearmodels.shared.linalg import has_constant
 from linearmodels.shared.rsquared import r_squared
@@ -38,6 +44,7 @@
 
 # The names of the covariance estimators of the models
 COV_TYPES = ("heteroskedastic", "kernel", "robust")
+_COV_KEYS = {"heteroskedastic": (), "robust": (), "kernel": ("kernel", "bandwidth")}
 
 
 def callback_factory(
@@ -422,7 +429,7 @@ def fit(
         Parameters
         ----------
         cov_type : {"robust", "heteroskedastic", "kernel"}, optional
-            Name of covariance estimator. The names are case sensitive. The
+            Name of covariance estimator. The names are not case sensitive. The
             default is "robust". "heteroskedastic" is the same as "robust".
         debiased : bool, optional
             Flag indicating whether to debias the covariance estimator using
@@ -446,6 +453,11 @@ def fit(
             If ``debiased`` is True and the number of observations is not larger
             than the number of factors plus one, or if the bandwidth is selected
             from the data and there are too few observations. See Notes.
+        ValueError
+            If a key of ``cov_config`` is not accepted by the covariance estimator
+            that ``cov_type`` selects. The message names the key and lists the
+            accepted keys. It is raised before the model is estimated. Releases
+            before 7.1 ignored the key.
 
         Notes
         -----
@@ -492,6 +504,8 @@ def fit(
            alphas and the loadings, ``risk_premia_se`` and the covariance with
            ``debiased=False`` do not change.
         """
+        cov_type = lower_option(cov_type)
+        check_cov_config(cov_config, cov_type, _COV_KEYS, KERNEL_LOOKUP)
         self._check_debiased(debiased)
         p = self.portfolios.ndarray
         f = self.factors.ndarray
@@ -933,7 +947,7 @@ def fit(
         Parameters
         ----------
         cov_type : {"robust", "heteroskedastic", "kernel"}, optional
-            Name of covariance estimator. The names are case sensitive. The
+            Name of covariance estimator. The names are not case sensitive. The
             default is "robust". "heteroskedastic" is the same as "robust".
         debiased : bool, optional
             Flag indicating whether to debias the covariance estimator using
@@ -957,6 +971,11 @@ def fit(
             If ``debiased`` is True and the number of observations is not larger
             than the number of factors plus one, or if the bandwidth is selected
             from the data and there are too few observations. See Notes.
+        ValueError
+            If a key of ``cov_config`` is not accepted by the covariance estimator
+            that ``cov_type`` selects. The message names the key and lists the
+            accepted keys. It is raised before the model is estimated. Releases
+            before 7.1 ignored the key.
 
         Notes
         -----
@@ -972,6 +991,8 @@ def fit(
         the kernel with fewer observations. See
         :func:`~linearmodels.iv.covariance.kernel_optimal_bandwidth`.
         """
+        cov_type = lower_option(cov_type)
+        check_cov_config(cov_config, cov_type, _COV_KEYS, KERNEL_LOOKUP)
         nobs, nf, nport, _, s1, s2, s3 = self._boundaries()
         excess_returns = not self._risk_free
         self._check_debiased(debiased)
@@ -1372,7 +1393,7 @@ def fit(
             Maximum number of iterations when minimizing objective. Must be
             positive. The default is 1000.
         cov_type : {"robust", "heteroskedastic", "kernel"}, optional
-            Name of covariance estimator. The names are case sensitive. The
+            Name of covariance estimator. The names are not case sensitive. The
             default is "robust". "heteroskedastic" is the same as "robust".
         debiased : bool, optional
             Flag indicating whether to debias the covariance estimator using
@@ -1404,6 +1425,11 @@ def fit(
             If ``starting`` does not have one element for each parameter, or if
             the bandwidth is selected from the data and there are too few
             observations. See Notes.
+        ValueError
+            If a key of ``cov_config`` is not accepted by the covariance estimator
+            that ``cov_type`` selects. The message names the key and lists the
+            accepted keys. It is raised before the model is estimated. Releases
+            before 7.1 ignored the key.
 
         Notes
         -----
@@ -1420,6 +1446,8 @@ def fit(
         observations. See
         :func:`~linearmodels.iv.covariance.kernel_optimal_bandwidth`.
         """
+        cov_type = lower_option(cov_type)
+        check_cov_config(cov_config, cov_type, _COV_KEYS, KERNEL_LOOKUP)
 
         nobs, n = self.portfolios.shape
         k = self.factors.shape[1]

linearmodels/iv/absorbing.py

@@ -46,8 +46,10 @@
 from scipy.sparse.linalg import lsmr
 
 from linearmodels.iv.common import f_statistic
+from linearmodels.iv.covariance import KERNEL_LOOKUP
 from linearmodels.iv.data import IVData
 from linearmodels.iv.model import (
+    _LS_COV_KEYS,
     COVARIANCE_ESTIMATORS,
     ClusteredCovariance,
     HeteroskedasticCovariance,
@@ -69,7 +71,11 @@
     reference_labels,
     row_labels,
 )
-from linearmodels.shared.exceptions import lookup_option, missing_warning
+from linearmodels.shared.exceptions import (
+    check_cov_config,
+    lookup_option,
+    missing_warning,
+)
 from linearmodels.shared.hypotheses import InvalidTestStatistic, WaldTestStatistic
 from linearmodels.shared.inputs import as_array_like
 from linearmodels.shared.rsquared import r_squared
@@ -1514,6 +1520,12 @@ def fit(
         UnknownOptionError
             If ``cov_type`` or ``kernel`` is not supported. It is both a
             ``KeyError`` and a ``ValueError``.
+        ValueError
+            If a key of ``cov_config`` is not accepted by the covariance estimator
+            that ``cov_type`` selects. The message names the key and lists the
+            accepted keys. It is raised before the model is estimated. Releases
+            before 7.1 raised the ``TypeError`` of the constructor of the covariance
+            estimator.
 
         Warns
         -----
@@ -1577,6 +1589,7 @@ def fit(
         changing only the covariance estimator is fast. If any of them differs,
         the variables are absorbed again using the new values.
         """
+        check_cov_config(cov_config, cov_type, _LS_COV_KEYS, KERNEL_LOOKUP)
         method = _check_method(method)
         if lsmr_options and absorb_options is None:
             # The options are options of LSMR, so they can not be passed to HDFE

linearmodels/iv/covariance.py

@@ -28,7 +28,7 @@
 from numpy.linalg import inv, pinv
 
 from linearmodels.shared.covariance import cov_cluster, cov_kernel
-from linearmodels.shared.exceptions import unknown_option_error
+from linearmodels.shared.exceptions import lower_option, unknown_option_error
 from linearmodels.shared.inputs import as_ndarray
 from linearmodels.shared.linalg import has_constant
 import linearmodels.typing
@@ -244,6 +244,7 @@ def _bandwidth_rule(kernel: str) -> tuple[int, float, float]:
     UnknownOptionError
         If kernel is not one of the supported names.
     """
+    kernel = lower_option(kernel)
     if kernel in ("bartlett", "newey-west"):
         return 1, 1.1447, 2 / 9
     if kernel in ("qs", "andrews", "quadratic-spectral"):
@@ -792,6 +793,7 @@ def __init__(
     ):
         super().__init__(x, y, z, params, debiased, kappa, extra_df)
         self._kernels = KERNEL_LOOKUP
+        kernel = lower_option(kernel)
         self._kernel = kernel
         self._bandwidth = bandwidth
         self._auto_bandwidth = False

linearmodels/iv/gmm.py

@@ -8,7 +8,7 @@
 from typing import Any
 import warnings
 
-from numpy import asarray, ndarray, unique
+from numpy import asarray, errstate, ndarray, unique
 from numpy.linalg import inv
 
 from linearmodels.iv.covariance import (
@@ -21,6 +21,7 @@
 from linearmodels.shared.exceptions import (
     DefaultBandwidthWarning,
     lookup_option,
+    lower_option,
     unknown_option_error,
 )
 import linearmodels.typing.data
@@ -315,8 +316,11 @@ def __init__(
         super().__init__(center, debiased)
         self._bandwidth = bandwidth
         self._orig_bandwidth = bandwidth
+        kernel = lower_option(kernel)
         self._kernel = kernel
         self._kernels = KERNEL_LOOKUP
+        if isinstance(kernel, str) and kernel not in KERNEL_LOOKUP:
+            raise unknown_option_error("kernel", kernel, KERNEL_LOOKUP)
         # Selecting the bandwidth from the data is the default since 7.1, and it
         # warns since the default was nobs - 2
         self._default_bandwidth = bandwidth is None and optimal_bw is None
@@ -344,7 +348,10 @@ def _scores(
 
     def _estimate_bandwidth(self, ze: linearmodels.typing.data.Float64Array) -> int:
         """The bandwidth that is estimated from the moment conditions."""
-        g = ze / ze.std(0)[None, :]
+        # A moment that does not vary has a standard deviation of 0, and the series is
+        # NaN. kernel_optimal_bandwidth raises, so there is no need to warn
+        with errstate(divide="ignore", invalid="ignore"):
+            g = ze / ze.std(0)[None, :]
         g = g.sum(1)
         return kernel_optimal_bandwidth(g, self._kernel)
 
@@ -631,6 +638,7 @@ def __init__(
         **cov_config: str | bool,
     ) -> None:
         super().__init__(x, y, z, params, debiased)
+        cov_type = lower_option(cov_type)
         self._cov_type = cov_type
         self._cov_config = cov_config
         self.w = w