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

Class DistributionStats

java.lang.Object

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

All Implemented Interfaces:

Serializable, NormalizerStats

public class DistributionStats
extends Object
implements NormalizerStats

Statistics about the normal distribution of values in data (means and standard deviations). Can be constructed incrementally by using the DynamicCustomOpsBuilder, which is useful for obtaining these statistics from an iterator. Can also load and save from files.

Author:

Ede Meijer

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	DistributionStats.Builder 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 and Description
DistributionStats(@NonNull INDArray mean, @NonNull INDArray std)

Method Summary

Modifier and Type	Method and Description
static DistributionStats	load(@NonNull File meanFile, @NonNull File stdFile) Load distribution statistics from the file system
void	save(@NonNull File meanFile, @NonNull File stdFile) Save distribution statistics to the file system

Methods inherited from class java.lang.Object

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

Constructor Detail

DistributionStats

public DistributionStats(@NonNull
                         @NonNull INDArray mean,
                         @NonNull
                         @NonNull INDArray std)

Parameters:

mean - row vector of means

std - row vector of standard deviations

Method Detail

load

public static DistributionStats load(@NonNull
                                     @NonNull File meanFile,
                                     @NonNull
                                     @NonNull File stdFile)
                              throws IOException

Load distribution statistics from the file system

Parameters:

meanFile - file containing the means

stdFile - file containing the standard deviations

Throws:

IOException

save

public void save(@NonNull
                 @NonNull File meanFile,
                 @NonNull
                 @NonNull File stdFile)
          throws IOException

Save distribution statistics to the file system

Parameters:

meanFile - file to contain the means

stdFile - file to contain the standard deviations

Throws:

IOException