com.numericalmethod.suanshu.stats.regression.linear.glm.quasi.family
Interface QuasiFamily

All Superinterfaces:

ExponentialDistribution

All Known Implementing Classes:

Binomial, Gamma, Gaussian, InverseGaussian, Poisson

public interface QuasiFamily
extends ExponentialDistribution

This interface represents the quasi family of GLM.

Method Summary

Family	asFamily()
LinkFunction	link()
double	quasiDeviance(double y, double mu) the quasi-deviance function corresponding to a single observation
double	quasiLikelihood(double mu, double y) the quasi-likelihood function corresponding to a single observation Q(μ; y)

Methods inherited from interface com.numericalmethod.suanshu.stats.regression.linear.glm.distribution.ExponentialDistribution

AIC, cumulant, deviance, dispersion, overdispersion, theta, variance

Method Detail

asFamily

Family asFamily()

link

LinkFunction link()

Returns:

the link function of this distribution

quasiLikelihood

double quasiLikelihood(double mu,
                       double y)

the quasi-likelihood function corresponding to a single observation Q(μ; y)

Parameters:

mu - μ

y - y

Returns:

Q(μ; y)

See Also:

"P. J. MacCullagh and J. A. Nelder, Generalized Linear Models, 2nd ed. Chapter 9. Table 9.1. p.326."

quasiDeviance

double quasiDeviance(double y,
                     double mu)

the quasi-deviance function corresponding to a single observation

Parameters:

y - y

mu - μ

Returns:

D(y; μ;)

See Also:

"P. J. MacCullagh and J. A. Nelder, Generalized Linear Models, 2nd ed. Chapter 9. Eq. 9.4., the integral form, p.327."