weka.dl4j

Class ShufflingDataSetIterator

java.lang.Object

weka.dl4j.ShufflingDataSetIterator

All Implemented Interfaces:

java.io.Serializable, java.util.Iterator<org.nd4j.linalg.dataset.DataSet>, org.nd4j.linalg.dataset.api.iterator.DataSetIterator

public class ShufflingDataSetIterator
extends java.lang.Object
implements org.nd4j.linalg.dataset.api.iterator.DataSetIterator, java.io.Serializable

An nd4j mini-batch iterator that shuffles the data whenever it is reset.

Version:

$Revision: 11711 $

Author:

Christopher Beckham, Eibe Frank

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
ShufflingDataSetIterator(org.nd4j.linalg.dataset.DataSet data, int batchSize, int seed) Constructs a new shuffling iterator.

Method Summary

Modifier and Type	Method and Description
boolean	asyncSupported() Whether the iterator can be used asynchronously.
int	batch() The size of the mini batches.
int	cursor() The cursor given the location in the dataset.
java.util.List<java.lang.String>	getLabels() Gets the labels.
org.nd4j.linalg.dataset.api.DataSetPreProcessor	getPreProcessor() Gets the preprocessor.
boolean	hasNext() Whether another batch of data is still available.
int	inputColumns() Returns the number of input columns.
org.nd4j.linalg.dataset.DataSet	next() Returns the next mini batch of data.
org.nd4j.linalg.dataset.DataSet	next(int num) Returns a batch of the given size
int	numExamples() The number of examples in the dataset.
void	remove() Enables removing of a mini-batch.
void	reset() Resets the cursor.
boolean	resetSupported() Whether the iterator can be reset.
void	setPreProcessor(org.nd4j.linalg.dataset.api.DataSetPreProcessor preProcessor) Sets the preprocessor.
int	totalExamples() Returns the total number of examples in the dataset.
int	totalOutcomes() Returns the total number of labels.

Methods inherited from class java.lang.Object

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

Methods inherited from interface java.util.Iterator

forEachRemaining

Constructor Detail

ShufflingDataSetIterator

public ShufflingDataSetIterator(org.nd4j.linalg.dataset.DataSet data,
                                int batchSize,
                                int seed)

Constructs a new shuffling iterator.

Parameters:

data - the dataset to operate on

batchSize - the mini batch size

seed - the seed for the random number generator

Method Detail

hasNext

public boolean hasNext()

Whether another batch of data is still available.

Specified by:

hasNext in interface java.util.Iterator<org.nd4j.linalg.dataset.DataSet>

Returns:

true if another batch is still available

next

public org.nd4j.linalg.dataset.DataSet next()

Returns the next mini batch of data.

Specified by:

next in interface java.util.Iterator<org.nd4j.linalg.dataset.DataSet>

Returns:

the dataset corresponding to the mini batch

next

public org.nd4j.linalg.dataset.DataSet next(int num)

Returns a batch of the given size

Specified by:

next in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Parameters:

num - the size of the batch to return

Returns:

a mini-batch of the given size

totalExamples

public int totalExamples()

Returns the total number of examples in the dataset.

Specified by:

totalExamples in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

the total number of examples in the dataset.

inputColumns

public int inputColumns()

Returns the number of input columns.

Specified by:

inputColumns in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

the number of input columns

totalOutcomes

public int totalOutcomes()

Returns the total number of labels.

Specified by:

totalOutcomes in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

the total number of labels

reset

public void reset()

Resets the cursor. Also shuffles the data again.

Specified by:

reset in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

resetSupported

public boolean resetSupported()

Whether the iterator can be reset.

Specified by:

resetSupported in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

true

asyncSupported

public boolean asyncSupported()

Whether the iterator can be used asynchronously.

Specified by:

asyncSupported in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

false

batch

public int batch()

The size of the mini batches.

Specified by:

batch in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

the size of the mini batches

cursor

public int cursor()

The cursor given the location in the dataset.

Specified by:

cursor in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

cursor

numExamples

public int numExamples()

The number of examples in the dataset.

Specified by:

numExamples in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

the number of examples

setPreProcessor

public void setPreProcessor(org.nd4j.linalg.dataset.api.DataSetPreProcessor preProcessor)

Sets the preprocessor.

Specified by:

setPreProcessor in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Parameters:

preProcessor -

getPreProcessor

public org.nd4j.linalg.dataset.api.DataSetPreProcessor getPreProcessor()

Gets the preprocessor.

Specified by:

getPreProcessor in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

preProcessor

getLabels

public java.util.List<java.lang.String> getLabels()

Gets the labels.

Specified by:

getLabels in interface org.nd4j.linalg.dataset.api.iterator.DataSetIterator

Returns:

the labels

remove

public void remove()

Enables removing of a mini-batch.

Specified by:

remove in interface java.util.Iterator<org.nd4j.linalg.dataset.DataSet>