linearmodels.panel.covariance.ClusteredCovariance¶

class linearmodels.panel.covariance.ClusteredCovariance(y: ndarray[tuple[int, ...], dtype[float64]], x: ndarray[tuple[int, ...], dtype[float64]], params: ndarray[tuple[int, ...], dtype[float64]], entity_ids: ndarray[tuple[int, ...], dtype[integer]], time_ids: ndarray[tuple[int, ...], dtype[integer]], *, debiased: bool = False, extra_df: int = 0, clusters: ndarray | DataArray | DataFrame | Series | list[Any] | tuple[Any, ...] | None = None, group_debias: bool = False)[source]¶

One-way (Rogers) or two-way clustered covariance estimation.

Parameters:¶

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

(entity x time) by 1 stacked array of the dependent variable.

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

(entity x time) by variables stacked array of the exogenous variables.

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

Variables by 1 array of estimated model parameters.

entity_ids: ndarray[tuple[int, ...], dtype[integer]]¶

(entity x time) by 1 stacked array of entity ids.

time_ids: ndarray[tuple[int, ...], dtype[integer]]¶

(entity x time) by 1 stacked array of time ids.

debiased: bool = False¶

Flag indicating whether to debias the estimator. The default is False.

extra_df: int = 0¶

Additional degrees of freedom consumed by models beyond the number of columns in x, e.g., fixed effects. Covariance estimators are always adjusted for extra_df irrespective of the setting of debiased. The default is 0.

clusters: ndarray | DataArray | DataFrame | Series | list[Any] | tuple[Any, ...] | None = None¶

Array of cluster group ids with nobs rows and 1 or 2 columns. If not provided, every observation is its own cluster.

group_debias: bool = False¶

Flag indicating whether to apply small-number of groups adjustment. The default is False.

Raises:¶

ValueError – If the model has no residual degrees of freedom: the number of observations less extra_df, and less the number of regressors if debiased is True, is not positive. It is also raised if clusters has more than 2 columns or does not have nobs rows.

Notes

The estimator of the covariance is

\[\hat{\Sigma}_{xx}^{-1}\hat{S}_{\mathcal{G}}\hat{\Sigma}_{xx}^{-1}\]

where

\[\hat{\Sigma}_{xx} = X'X\]

and \(\hat{S}_{\mathcal{G}}\) is a one- or two-way cluster covariance of the scores. Two-way clustering is implemented by summing up the two one-way cluster covariances and then subtracting the one-way clustering covariance computed using the group formed from the intersection of the two groups.

Two small sample adjustments are available. debiased=True will account for regressors in the main model. group_debias=True will provide a small sample adjustment for the number of clusters of the form

\[(g / (g- 1)) ((n - 1) / n)\]

where g is the number of distinct groups and n is the number of observations.

Methods

deferred_cov()

Covariance calculation deferred until executed.

Properties

ALLOWED_KWARGS
DEFAULT_KERNEL
cov	Estimated covariance.
eps	Model residuals.
name	Covariance estimator name.
s2	Error variance.