org.apache.spark.ml.optim.aggregator

Interface DifferentiableLossAggregator<Datum,Agg extends DifferentiableLossAggregator<Datum,Agg>>

Type Parameters:

Datum - The type of the instances added to the aggregator to update the loss and gradient.

Agg - Specialization of DifferentiableLossAggregator. Classes that subclass this type need to use this parameter to specify the concrete type of the aggregator.

All Superinterfaces:

java.io.Serializable

public interface DifferentiableLossAggregator<Datum,Agg extends DifferentiableLossAggregator<Datum,Agg>>
extends scala.Serializable

A parent trait for aggregators used in fitting MLlib models. This parent trait implements some of the common code shared between concrete instances of aggregators. Subclasses of this aggregator need only implement the add method.

Method Summary

Modifier and Type	Method and Description
Agg	add(Datum instance) Add a single data point to this aggregator.
int	dim() The dimension of the gradient array.
Vector	gradient() The current weighted averaged gradient.
double[]	gradientSumArray() Array of gradient values that are mutated when new instances are added to the aggregator.
double	loss() The current loss value of this aggregator.
double	lossSum()
Agg	merge(Agg other) Merge two aggregators.
double	weight() Weighted count of instances in this aggregator.
double	weightSum()

Method Detail

add

Agg add(Datum instance)

Add a single data point to this aggregator.

dim

int dim()

The dimension of the gradient array.

gradient

Vector gradient()

The current weighted averaged gradient.

gradientSumArray

double[] gradientSumArray()

Array of gradient values that are mutated when new instances are added to the aggregator.

loss

double loss()

The current loss value of this aggregator.

lossSum

double lossSum()

merge

Agg merge(Agg other)

Merge two aggregators. The `this` object will be modified in place and returned.

weight

double weight()

Weighted count of instances in this aggregator.

weightSum

double weightSum()