Binomial exponential family distribution.
Parameters: | link : a link instance, optional
|
---|
Notes
endog for Binomial can be specified in one of three ways.
Attributes
Binomial.link | a link instance | The link function of the Binomial instance |
Binomial.variance | varfunc instance | variance is an instance of statsmodels.family.varfuncs.binary |
Methods
deviance(endog, mu[, scale]) | Deviance function for either Bernoulli or Binomial data. |
fitted(lin_pred) | Fitted values based on linear predictors lin_pred. |
initialize(endog) | Initialize the response variable. |
loglike(endog, mu[, scale]) | Loglikelihood function for Binomial exponential family distribution. |
predict(mu) | Linear predictors based on given mu values. |
resid_anscombe(endog, mu) | The Anscombe residuals |
resid_dev(endog, mu[, scale]) | Binomial deviance residuals |
starting_mu(y) | The starting values for the IRLS algorithm for the Binomial family. |
weights(mu) | Weights for IRLS steps |