org.encog.ml.data.versatile.normalizers

Class IndexedNormalizer

java.lang.Object

org.encog.ml.data.versatile.normalizers.IndexedNormalizer

All Implemented Interfaces:

Normalizer

public class IndexedNormalizer
extends Object
implements Normalizer

Normalize ordinal/nominal values to a single value that is simply the index of the class in the list. For example, "one", "two", "three" normalizes to 0,1,2.

Constructor Summary

Constructors

Constructor and Description
IndexedNormalizer()

Method Summary

Methods

Modifier and Type	Method and Description
String	denormalizeColumn(ColumnDefinition colDef, MLData data, int dataColumn) Denormalize a value.
int	normalizeColumn(ColumnDefinition colDef, double value, double[] outputData, int outputColumn) Normalize a column from a double.
int	normalizeColumn(ColumnDefinition colDef, String value, double[] outputData, int outputColumn) Normalize a column from a string.
int	outputSize(ColumnDefinition colDef) Determine the normalized size of the specified column.

Methods inherited from class java.lang.Object

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

Constructor Detail

IndexedNormalizer

public IndexedNormalizer()

Method Detail

outputSize

public int outputSize(ColumnDefinition colDef)

Determine the normalized size of the specified column.

Specified by:

outputSize in interface Normalizer

Parameters:

colDef - The column to check.

Returns:

The size of the column normalized.

normalizeColumn

public int normalizeColumn(ColumnDefinition colDef,
                  String value,
                  double[] outputData,
                  int outputColumn)

Normalize a column from a string. The output will go to an array, starting at outputColumn.

Specified by:

normalizeColumn in interface Normalizer

Parameters:

colDef - The column that is being normalized.

value - The value to normalize.

outputData - The array to output to.

outputColumn - The index to start at in outputData.

Returns:

The new index (in outputData) that we've moved to.

denormalizeColumn

public String denormalizeColumn(ColumnDefinition colDef,
                       MLData data,
                       int dataColumn)

Denormalize a value.

Specified by:

denormalizeColumn in interface Normalizer

Parameters:

colDef - The column to denormalize.

data - The data to denormalize.

dataColumn - The starting location inside data.

Returns:

The denormalized value.

normalizeColumn

public int normalizeColumn(ColumnDefinition colDef,
                  double value,
                  double[] outputData,
                  int outputColumn)

Normalize a column from a double. The output will go to an array, starting at outputColumn.

Specified by:

normalizeColumn in interface Normalizer

Parameters:

colDef - The column that is being normalized.

value - The value to normalize.

outputData - The array to output to.

outputColumn - The index to start at in outputData.

Returns:

The new index (in outputData) that we've moved to.