#include <genesis/utils/math/regression/family.hpp>

Detailed Description

Definition at line 48 of file family.hpp.

Public Types
enum	Family { kUnknown, kBinomial, kPoisson, kGaussian, kGamma }
	List of common GLM Families. More...

Public Attributes
std::function< GlmLink()>	canonical_link
	Get the canonical link function. More...

GlmLink::Link	canonical_link_id = GlmLink::kUnknown
	Internal ID of the GlmLink, used to check if the link is the canonical one for a given distribution family. More...

Family	id = kUnknown
	Internal ID of the GlmFamily, used to check for specific families where needed. More...

std::function< double(double y, double mu)>	log_likelihood
	Log-Likelihood contribution of a value. To be multiplied by prior weight. More...

std::function< double(double mu)>	rectify
	Rectify to a valid value, for the fitted mean, to avoid extreme predictions. More...

std::function< double(double y, double mu)>	unit_deviance
	Unit deviance for the distribution family. More...

std::function< double(double mu)>	variance
	Variance function for the distribution family. More...

Member Enumeration Documentation

◆ Family

enum Family

List of common GLM Families.

Enumerator
kUnknown
kBinomial
kPoisson
kGaussian
kGamma

Definition at line 53 of file family.hpp.

Member Data Documentation

◆ canonical_link

std::function<GlmLink()> canonical_link

Get the canonical link function.

Definition at line 95 of file family.hpp.

◆ canonical_link_id

GlmLink::Link canonical_link_id = GlmLink::kUnknown

Internal ID of the GlmLink, used to check if the link is the canonical one for a given distribution family.

Definition at line 70 of file family.hpp.

◆ id

Family id = kUnknown

Internal ID of the GlmFamily, used to check for specific families where needed.

Definition at line 64 of file family.hpp.

◆ log_likelihood

std::function<double( double y, double mu )> log_likelihood

Log-Likelihood contribution of a value. To be multiplied by prior weight.

Definition at line 80 of file family.hpp.

◆ rectify

std::function<double( double mu )> rectify

Rectify to a valid value, for the fitted mean, to avoid extreme predictions.

Definition at line 85 of file family.hpp.

◆ unit_deviance

std::function<double( double y, double mu )> unit_deviance

Unit deviance for the distribution family.

Definition at line 90 of file family.hpp.

◆ variance

std::function<double( double mu )> variance

Variance function for the distribution family.

Definition at line 75 of file family.hpp.

The documentation for this struct was generated from the following file:

family.hpp