org.apache.spark.ml.regression

Class GeneralizedLinearRegression.Poisson$

Object

org.apache.spark.ml.regression.GeneralizedLinearRegression.Poisson$

All Implemented Interfaces:

java.io.Serializable

Enclosing class:

GeneralizedLinearRegression

public static class GeneralizedLinearRegression.Poisson$
extends Object

Poisson exponential family distribution. The default link for the Poisson family is the log link.

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static GeneralizedLinearRegression.Poisson$	MODULE$ Static reference to the singleton instance of this Scala object.

Constructor Summary

Constructors

Constructor and Description
Poisson$()

Method Summary

Modifier and Type	Method and Description
double	aic(RDD<scala.Tuple3<Object,Object,Object>> predictions, double deviance, double numInstances, double weightSum) Akaike Information Criterion (AIC) value of the family for a given dataset.
org.apache.spark.ml.regression.GeneralizedLinearRegression.Link	defaultLink() The default link instance of this family.
double	deviance(double y, double mu, double weight) Deviance of (y, mu) pair.
double	initialize(double y, double weight) Initialize the starting value for mu.
String	name()
double	project(double mu) Trim the fitted value so that it will be in valid range.
double	variance(double mu) The variance of the endogenous variable's mean, given the value mu.
double	variancePower()

Methods inherited from class Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

MODULE$

public static final GeneralizedLinearRegression.Poisson$ MODULE$

Static reference to the singleton instance of this Scala object.

Constructor Detail

Poisson$

public Poisson$()

Method Detail

name

public String name()

defaultLink

public org.apache.spark.ml.regression.GeneralizedLinearRegression.Link defaultLink()

The default link instance of this family.

initialize

public double initialize(double y,
                         double weight)

Initialize the starting value for mu.

variance

public double variance(double mu)

The variance of the endogenous variable's mean, given the value mu.

deviance

public double deviance(double y,
                       double mu,
                       double weight)

Deviance of (y, mu) pair.

aic

public double aic(RDD<scala.Tuple3<Object,Object,Object>> predictions,
                  double deviance,
                  double numInstances,
                  double weightSum)

Akaike Information Criterion (AIC) value of the family for a given dataset.

Parameters:

predictions - an RDD of (y, mu, weight) of instances in evaluation dataset

deviance - the deviance for the fitted model in evaluation dataset

numInstances - number of instances in evaluation dataset

weightSum - weights sum of instances in evaluation dataset

Returns:

(undocumented)

variancePower

public double variancePower()

project

public double project(double mu)

Trim the fitted value so that it will be in valid range.