Package org.nd4j.linalg.dataset.api.preprocessor.stats

Class DistributionStats.Builder

java.lang.Object

org.nd4j.linalg.dataset.api.preprocessor.stats.DistributionStats.Builder

All Implemented Interfaces:

NormalizerStats.Builder<DistributionStats>

Enclosing class:

DistributionStats

public static class DistributionStats.Builder
extends Object
implements NormalizerStats.Builder<DistributionStats>

DynamicCustomOpsBuilder class that can incrementally update a running mean and variance in order to create statistics for a large set of data

Constructor Summary

Constructors

Constructor	Description
Builder()

Method Summary

Modifier and Type	Method	Description
DistributionStats.Builder	add(@NonNull INDArray data, INDArray mask)	Add rows of data to the statistics
DistributionStats.Builder	addFeatures(DataSet dataSet)	Add the features of a DataSet to the statistics
DistributionStats.Builder	addLabels(DataSet dataSet)	Add the labels of a DataSet to the statistics
DistributionStats	build()	Create a DistributionStats object from the data ingested so far.

Methods inherited from class java.lang.Object

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

Constructor Detail

Builder

public Builder()

Method Detail

addFeatures

public DistributionStats.Builder addFeatures(@NonNull
                                             DataSet dataSet)

Add the features of a DataSet to the statistics

Specified by:

addFeatures in interface NormalizerStats.Builder<DistributionStats>

addLabels

public DistributionStats.Builder addLabels(@NonNull
                                           DataSet dataSet)

Add the labels of a DataSet to the statistics

Specified by:

addLabels in interface NormalizerStats.Builder<DistributionStats>

add

public DistributionStats.Builder add(@NonNull
                                     @NonNull INDArray data,
                                     INDArray mask)

Add rows of data to the statistics

Specified by:

add in interface NormalizerStats.Builder<DistributionStats>

Parameters:

data - the matrix containing multiple rows of data to include

mask - (optionally) the mask of the data, useful for e.g. time series

build

public DistributionStats build()

Create a DistributionStats object from the data ingested so far. Can be used multiple times when updating online.

Specified by:

build in interface NormalizerStats.Builder<DistributionStats>

Returns: