statsmodels.genmod.generalized_linear_model.GLM.information#

GLM.information(params, scale=None)[source]#

Fisher information matrix.

Returns -1 * Hessian of the log-likelihood evaluated at params, i.e. the expected information matrix, which is positive definite.

Parameters:

paramsndarray: Parameter at which the information is evaluated.
scalefloat, optional: If scale is None, then the default scale will be calculated. Default scale is defined by self.scaletype and set in fit. If scale is not None, then it is used as a fixed scale.

Returns:

ndarray: The expected information matrix, X' W X / scale. This is -hessian(params, scale, observed=False). It differs from -hessian(params, scale), which is the observed information by default, unless the link is the canonical link of the family.