statsmodels.genmod.generalized_linear_model.GLM.estimate_tweedie_power

GLM.estimate_tweedie_power(mu, method='brentq', low=1.01, high=5.0)[source]

Tweedie specific function to estimate scale and the variance parameter. The variance parameter is also referred to as p, xi, or shape.

Parameters:
  • mu (array-like) – Fitted mean response variable
  • method (str, defaults to 'brentq') – Scipy optimizer used to solve the Pearson equation. Only brentq currently supported.
  • low (float, optional) – Low end of the bracketing interval [a,b] to be used in the search for the power. Defaults to 1.01.
  • high (float, optional) – High end of the bracketing interval [a,b] to be used in the search for the power. Defaults to 5.
Returns:

power – The estimated shape or power
Return type:

float