statsmodels.duration.hazard_regression.PHReg.robust_covariance

PHReg.robust_covariance(params) [source]

Returns a covariance matrix for the proportional hazards model regresion coefficient estimates that is robust to certain forms of model misspecification.

Parameters:	params (ndarray) – The parameter vector at which the covariance matrix is calculated.
Returns:
Return type:	The robust covariance matrix as a square ndarray.

Notes

This function uses the groups argument to determine groups within which observations may be dependent. The covariance matrix is calculated using the Huber-White “sandwich” approach.