linearmodels.asset_pricing.covariance.KernelCovariance¶

class linearmodels.asset_pricing.covariance.KernelCovariance(xe: ndarray[tuple[int, ...], dtype[float64]], *, jacobian: ndarray | None = None, inv_jacobian: ndarray | None = None, kernel: str | None = None, bandwidth: float | None = None, center: bool = True, debiased: bool = False, df: int = 0)[source]¶

Heteroskedasticity-autocorrelation (HAC) robust covariance estimator.

Parameters:¶

xe: ndarray[tuple[int, ...], dtype[float64]]¶

The scores (moment) conditions, with shape (nobs, nmoments).

jacobian: ndarray | None = None¶

Jacobian, with shape (nmoments, nparams). One and only one of jacobian and inv_jacobian must be provided.

inv_jacobian: ndarray | None = None¶

Inverse jacobian. One and only one of jacobian and inv_jacobian must be provided.

kernel: str | None = None¶

Name of the kernel. The default is “bartlett”. The names are not case sensitive. Supported kernels are:

”bartlett”, “newey-west” - Bartlett’s kernel
”parzen”, “gallant” - Parzen’s kernel
”qs”, “quadratic-spectral”, “andrews” - Quadratic spectral kernel

bandwidth: float | None = None¶

Non-negative bandwidth. If None, the optimal bandwidth is estimated. It needs at least 4 observations with the Bartlett kernel, 5 with the Parzen kernel and 6 with the quadratic spectral kernel, and a ValueError is raised for fewer when the bandwidth is first used. See kernel_optimal_bandwidth().

center: bool = True¶

Flag indicating to center the scores when computing the covariance. The default is True.

debiased: bool = False¶

Flag indicating to use a debiased estimator. The default is False.

df: int = 0¶

Degree of freedom value to use if debiasing. The default is 0.

Raises:¶

ValueError – If both or neither of jacobian and inv_jacobian are provided, or if bandwidth is negative.
TypeError – If kernel is not a str, or bandwidth is not a number.
UnknownOptionError – If kernel is not supported. It is both a KeyError and a ValueError.

See also

linearmodels.iv.covariance.kernel_weight_bartlett: Weights of the Bartlett kernel.
linearmodels.iv.covariance.kernel_weight_parzen: Weights of the Parzen kernel.
linearmodels.iv.covariance.kernel_weight_quadratic_spectral: Weights of the quadratic spectral kernel.

Notes

Changed in version 7.1: The optimal bandwidth raises a ValueError if there are fewer observations than the kernel needs, and if it is not finite. Earlier versions raised “cannot convert float infinity to integer” (an OverflowError) or “cannot convert float NaN to integer” for some data, and used a bandwidth of nobs - 1, or another bandwidth, for other data.

Methods

Properties

`bandwidth`	Bandwidth used in estimation.
`config`	Configuration of the covariance estimator.
`cov`	Compute parameter covariance.
`inv_jacobian`	Inverse Jacobian.
`jacobian`	The Jacobian.
`kernel`	Kernel used in estimation.
`s`	Score/moment condition covariance.
`square`	Flag indicating if jacobian is square.