linearmodels.asset_pricing.covariance.KernelCovariance¶
- class KernelCovariance(xe, *, jacobian=None, inv_jacobian=None, kernel=None, bandwidth=None, center=True, debiased=False, df=0)[source]¶
Heteroskedasticity-autocorrelation (HAC) robust covariance estimator
- Parameters
- xe
ndarray
The scores (moment) conditions.
- jacobian
ndarray
Jacobian. One and only one of jacobian and inv_jacobian must be provided.
- inv_jacobian
ndarray
Inverse jacobian. One and only one of jacobian and inv_jacobian must be provided.
- kernel
str
Kernel name. See notes for available kernels. The default is “bartlett”.
- bandwidth
int
Non-negative integer bandwidth. If None, the optimal bandwidth is estimated.
- centerbool
Flag indicating to center the scores when computing the covariance.
- debiasedbool
Flag indicating to use a debiased estimator.
- df
int
Degree of freedom value ot use if debiasing.
- xe
See also
- Attributes
Methods
Properties
Bandwidth used in estimation
Compute parameter covariance
Inverse Jacobian
The Jacobian
Kernel used in estimation
Score/moment condition covariance
Flag indicating if jacobian is square