Tweedie family.
Parameters: | link : a link instance, optional
var_power : float, optional
link_power : float, optional
|
---|
Notes
Logliklihood function not implemented because of the complexity of calculating an infinite series of summations. The variance power can be estimated using the estimate_tweedie_power function that is part of the GLM class.
Attributes
Tweedie.link | a link instance | The link function of the Tweedie instance |
Tweedie.variance | varfunc instance | variance is an instance of statsmodels.family.varfuncs.Power |
Tweedie.link_power | float | The power of the link function, or 0 if its a log link. |
Tweedie.var_power | float | The power of the variance function. |
Methods
deviance(endog, mu[, freq_weights, scale]) | Returns the value of the deviance function. |
fitted(lin_pred) | Fitted values based on linear predictors lin_pred. |
loglike(endog, mu[, freq_weights, scale]) | The log-likelihood function in terms of the fitted mean response. |
predict(mu) | Linear predictors based on given mu values. |
resid_anscombe(endog, mu) | The Anscombe residuals for the Tweedie family |
resid_dev(endog, mu[, scale]) | Tweedie Deviance Residual |
starting_mu(y) | Starting value for mu in the IRLS algorithm. |
variance | |
weights(mu) | Weights for IRLS steps |
Methods
deviance(endog, mu[, freq_weights, scale]) | Returns the value of the deviance function. |
fitted(lin_pred) | Fitted values based on linear predictors lin_pred. |
loglike(endog, mu[, freq_weights, scale]) | The log-likelihood function in terms of the fitted mean response. |
predict(mu) | Linear predictors based on given mu values. |
resid_anscombe(endog, mu) | The Anscombe residuals for the Tweedie family |
resid_dev(endog, mu[, scale]) | Tweedie Deviance Residual |
starting_mu(y) | Starting value for mu in the IRLS algorithm. |
weights(mu) | Weights for IRLS steps |