org.tribuo.classification.sgd

Class Util

java.lang.Object

org.tribuo.classification.sgd.Util

public class Util
extends Object

SGD utilities. Currently stores methods for shuffling examples and their associated labels and weights.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	Util.ExampleArray A nominal tuple.
static class	Util.SequenceExampleArray A nominal tuple.

Constructor Summary

Constructors

Constructor and Description
Util()

Method Summary

Modifier and Type	Method and Description
static Util.SequenceExampleArray	shuffle(SGDVector[][] features, int[][] labels, double[] weights, SplittableRandom rng) Shuffles a sequence of features, labels and weights, returning a tuple of the shuffled values.
static Util.ExampleArray	shuffle(SparseVector[] features, int[] labels, double[] weights, SplittableRandom rng) Shuffles the features, labels and weights returning a tuple of the shuffled inputs.
static void	shuffleInPlace(SGDVector[][] features, int[][] labels, double[] weights, SplittableRandom rng) In place shuffle used for sequence problems.
static void	shuffleInPlace(SparseVector[] features, int[] labels, double[] weights, int[] indices, SplittableRandom rng) In place shuffle of the features, labels, weights and indices.
static void	shuffleInPlace(SparseVector[] features, int[] labels, double[] weights, SplittableRandom rng) Deprecated. In favour of AbstractSGDTrainer.shuffleInPlace(org.tribuo.math.la.SGDVector[], T[], double[], java.util.SplittableRandom).

Methods inherited from class java.lang.Object

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

Constructor Detail

Util

public Util()

Method Detail

shuffleInPlace

@Deprecated
public static void shuffleInPlace(SparseVector[] features,
                                              int[] labels,
                                              double[] weights,
                                              SplittableRandom rng)

Deprecated. In favour of AbstractSGDTrainer.shuffleInPlace(org.tribuo.math.la.SGDVector[], T[], double[], java.util.SplittableRandom).

In place shuffle of the features, labels and weights.

Parameters:

features - Input features.

labels - Input labels.

weights - Input weights.

rng - SplittableRandom number generator.

shuffleInPlace

public static void shuffleInPlace(SparseVector[] features,
                                  int[] labels,
                                  double[] weights,
                                  int[] indices,
                                  SplittableRandom rng)

In place shuffle of the features, labels, weights and indices.

Parameters:

features - Input features.

labels - Input labels.

weights - Input weights.

indices - Input indices.

rng - SplittableRandom number generator.

shuffle

public static Util.ExampleArray shuffle(SparseVector[] features,
                                        int[] labels,
                                        double[] weights,
                                        SplittableRandom rng)

Shuffles the features, labels and weights returning a tuple of the shuffled inputs.

Parameters:

features - Input features.

labels - Input labels.

weights - Input weights.

rng - SplittableRandom number generator.

Returns:

A tuple of shuffled features, labels and weights.

shuffleInPlace

public static void shuffleInPlace(SGDVector[][] features,
                                  int[][] labels,
                                  double[] weights,
                                  SplittableRandom rng)

In place shuffle used for sequence problems.

Parameters:

features - Input features.

labels - Input labels.

weights - Input weights.

rng - SplittableRandom number generator.

shuffle

public static Util.SequenceExampleArray shuffle(SGDVector[][] features,
                                                int[][] labels,
                                                double[] weights,
                                                SplittableRandom rng)

Shuffles a sequence of features, labels and weights, returning a tuple of the shuffled values.

Parameters:

features - Input features.

labels - Input labels.

weights - Input weights.

rng - SplittableRandom number generator.

Returns:

A tuple of shuffled features, labels and weights.