Catalano.MachineLearning.Dataset

Class DatasetClassification

java.lang.Object

Catalano.MachineLearning.Dataset.DatasetClassification

All Implemented Interfaces:

IDataset<double[][],int[]>, java.io.Serializable

public class DatasetClassification
extends java.lang.Object
implements IDataset<double[][],int[]>

Dataset for classification.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
DatasetClassification(java.lang.String filepath) Initializes a new instance of the DatasetClassification class.
DatasetClassification(java.lang.String name, double[][] input, int[] output) Initializes a new instance of the DatasetClassification class.
DatasetClassification(java.lang.String name, double[][] input, int[] output, DecisionVariable[] attributes) Initializes a new instance of the DatasetClassification class.
DatasetClassification(java.lang.String name, double[][] input, int[] output, DecisionVariable[] attributes, int classIndex) Initializes a new instance of the DatasetClassification class.
DatasetClassification(java.lang.String filepath, java.lang.String name) Initializes a new instance of the DatasetClassification class.
DatasetClassification(java.lang.String filepath, java.lang.String name, boolean ignoreAttributeInfo) Initializes a new instance of the DatasetClassification class.
DatasetClassification(java.lang.String filepath, java.lang.String name, boolean ignoreAttributeInfo, int classIndex) Initializes a new instance of the DatasetClassification class.

Method Summary

Modifier and Type	Method and Description
static DatasetClassification	FromCSV(java.lang.String filepath, java.lang.String name) Read dataset from a CSV structure.
static DatasetClassification	FromCSV(java.lang.String filepath, java.lang.String name, boolean ignoreAttributeInfo) Read dataset from a CSV structure.
static DatasetClassification	FromCSV(java.lang.String filepath, java.lang.String name, boolean ignoreAttributeInfo, int classIndex) Construct a dataset from an CSV file.
int	getClassIndex() Get the class index.
Codebook	getCodebook() Get Codebook.
DecisionVariable[]	getDecisionVariables() Get the decision variables.
IFeatureScaling	getFeatureScaling() Get normalization.
double[][]	getInput() Get the input data.
double[][]	getInput(int label) Get all the instances with the related label.
java.lang.String	getName() Get the name of the dataset.
int	getNumberOfAttributes() Get the number of all the attributes.
int	getNumberOfClasses() Get the number of classes.
int	getNumberOfContinuous() Get the number of type continuous.
int	getNumberOfDiscrete() Get the number of type discrete.
int	getNumberOfInstances() Get the number of instances.
int[]	getOutput() Get the output data.
StatisticsDataset[]	getStatistics() Get the statistics from the dataset.
void	Imputation(IImputation filter) Perform imputation algorithms.
void	KeepAttributes(int[] indexes) Keep the selected attributes and remove the rest.
void	Normalize() Normalize all continuous data.
void	Normalize(double min, double max) Normalize all continuous data.
void	Normalize(IFeatureScaling normalization) Normalize all continuous data.
void	RemoveAttribute(int index) Remove an attribute.
void	RemoveAttribute(int[] indexes) Remove attributes.
void	RemoveClass(int id) Remove all instances within related class.
void	RemoveMissingInstances() Remove all the instances where is missing values.
void	setClassIndex(int classIndex) Set the class index.
void	setInput(double[][] input, DecisionVariable[] variables) Set the input data.
void	Shuffle() Shuffle the dataset.
void	Shuffle(long seed) Shuffle the dataset.
void	Sort() Sort the dataset in relation of the labels.
DatasetClassification	Split(float percentage) Split the percentange of the dataset in Trainning and the rest in Validation.
DatasetClassification	Split(float percentage, java.lang.String name) Split the percentange of the dataset in Trainning and the rest in Validation.
void	Standartize() Standartize all continuous data.
void	WriteAsARFF(java.lang.String filename) Write dataset as ARFF file.
void	WriteAsARFF(java.lang.String filename, int decimalPlaces) Write dataset as ARFF file.
void	WriteAsCSV(java.lang.String filename) Write the dataset as CSV file.
void	WriteAsCSV(java.lang.String filename, int decimalPlaces) Write the dataset as CSV file.
void	WriteAsCSV(java.lang.String filename, int decimalPlaces, char delimiter) Write the dataset as CSV file.
void	WriteAsCSV(java.lang.String filename, int decimalPlaces, char delimiter, java.lang.String newLine)
void	WriteAsCSV(java.lang.String filename, int decimalPlaces, char delimiter, java.lang.String newLine, boolean writeHeader) Write the dataset as CSV file.

Methods inherited from class java.lang.Object

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

Constructor Detail

DatasetClassification

public DatasetClassification(java.lang.String filepath)

Initializes a new instance of the DatasetClassification class.

Parameters:

filepath - File path.

DatasetClassification

public DatasetClassification(java.lang.String filepath,
                             java.lang.String name)

Initializes a new instance of the DatasetClassification class.

Parameters:

filepath - File path.

name - Name of the dataset.

DatasetClassification

public DatasetClassification(java.lang.String filepath,
                             java.lang.String name,
                             boolean ignoreAttributeInfo)

Initializes a new instance of the DatasetClassification class.

Parameters:

filepath - File path.

name - Name of the dataset.

ignoreAttributeInfo - Ignore first line (Attribute information).

DatasetClassification

public DatasetClassification(java.lang.String filepath,
                             java.lang.String name,
                             boolean ignoreAttributeInfo,
                             int classIndex)

Initializes a new instance of the DatasetClassification class.

Parameters:

filepath - File path.

name - Name of the dataset.

classIndex - Class index.

ignoreAttributeInfo - Ignore first line (Attribute information).

DatasetClassification

public DatasetClassification(java.lang.String name,
                             double[][] input,
                             int[] output)

Initializes a new instance of the DatasetClassification class.

Parameters:

name - Name of the dataset.

input - Input data.

output - Output data.

DatasetClassification

public DatasetClassification(java.lang.String name,
                             double[][] input,
                             int[] output,
                             DecisionVariable[] attributes)

Initializes a new instance of the DatasetClassification class.

Parameters:

name - Name of the dataset.

attributes - Decision variables.

input - Input data.

output - Output data.

DatasetClassification

public DatasetClassification(java.lang.String name,
                             double[][] input,
                             int[] output,
                             DecisionVariable[] attributes,
                             int classIndex)

Initializes a new instance of the DatasetClassification class.

Parameters:

name - Name of the dataset.

attributes - Decision variables.

input - Input data.

output - Output data.

classIndex - Class index.

Method Detail

getName

public java.lang.String getName()

Get the name of the dataset.

Returns:

Name.

getInput

public double[][] getInput()

Get the input data.

Specified by:

getInput in interface IDataset<double[][],int[]>

Returns:

Input data.

setInput

public void setInput(double[][] input,
                     DecisionVariable[] variables)

Description copied from interface: IDataset

Set the input data.

Specified by:

setInput in interface IDataset<double[][],int[]>

Parameters:

input - Input.

getOutput

public int[] getOutput()

Get the output data.

Specified by:

getOutput in interface IDataset<double[][],int[]>

Returns:

Output data.

getClassIndex

public int getClassIndex()

Get the class index.

Returns:

Class index.

setClassIndex

public void setClassIndex(int classIndex)

Set the class index.

Parameters:

classIndex - Class index.

getDecisionVariables

public DecisionVariable[] getDecisionVariables()

Get the decision variables.

Specified by:

getDecisionVariables in interface IDataset<double[][],int[]>

Returns:

Decision Variables.

getNumberOfInstances

public int getNumberOfInstances()

Get the number of instances.

Returns:

Number of instances.

getNumberOfAttributes

public int getNumberOfAttributes()

Get the number of all the attributes. Number of feature + class.

Returns:

Number of attributes.

getNumberOfClasses

public int getNumberOfClasses()

Get the number of classes.

Returns:

Number of classes.

getNumberOfContinuous

public int getNumberOfContinuous()

Get the number of type continuous.

Returns:

Number of type continuous.

getNumberOfDiscrete

public int getNumberOfDiscrete()

Get the number of type discrete.

Returns:

Number of type discrete.

getFeatureScaling

public IFeatureScaling getFeatureScaling()

Get normalization.

Returns:

Normalization.

getCodebook

public Codebook getCodebook()

Get Codebook.

Returns:

Codebook.

FromCSV

public static DatasetClassification FromCSV(java.lang.String filepath,
                                            java.lang.String name)

Read dataset from a CSV structure. The last column of CSV is interpreted as output.

Parameters:

filepath - File path.

name - Name of the dataset.

Returns:

Classification Dataset.

FromCSV

public static DatasetClassification FromCSV(java.lang.String filepath,
                                            java.lang.String name,
                                            boolean ignoreAttributeInfo)

Read dataset from a CSV structure. The last column of CSV is interpreted as output.

Parameters:

filepath - File path.

name - Name of the dataset.

ignoreAttributeInfo - Ignore attribute information.

Returns:

Classification dataset.

FromCSV

public static DatasetClassification FromCSV(java.lang.String filepath,
                                            java.lang.String name,
                                            boolean ignoreAttributeInfo,
                                            int classIndex)

Construct a dataset from an CSV file.

Parameters:

filepath - File.

name - Name of the dataset.

classIndex - Index of the attribute for to be setup as output.

ignoreAttributeInfo - Ignore attribute information.

Returns:

Classification Dataset.

getInput

public double[][] getInput(int label)

Get all the instances with the related label.

Parameters:

label - Label.

Returns:

Data.

Imputation

public void Imputation(IImputation filter)

Perform imputation algorithms.

Parameters:

filter - Filter.

Normalize

public void Normalize()

Normalize all continuous data. Default: (0..1).

Normalize

public void Normalize(double min,
                      double max)

Normalize all continuous data.

Parameters:

min - Minimum value.

max - Maximum value.

Normalize

public void Normalize(IFeatureScaling normalization)

Normalize all continuous data.

Parameters:

normalization - Normalization.

RemoveAttribute

public void RemoveAttribute(int index)

Remove an attribute.

Parameters:

index - Index of the attribute.

RemoveAttribute

public void RemoveAttribute(int[] indexes)

Remove attributes.

Parameters:

indexes - Indexes of the attributes.

KeepAttributes

public void KeepAttributes(int[] indexes)

Keep the selected attributes and remove the rest.

Parameters:

indexes - Indexes of the attributes.

RemoveClass

public void RemoveClass(int id)

Remove all instances within related class. The labels will be reindexed.

Parameters:

id - Class.

RemoveMissingInstances

public void RemoveMissingInstances()

Remove all the instances where is missing values.

Split

public DatasetClassification Split(float percentage)

Split the percentange of the dataset in Trainning and the rest in Validation.

Parameters:

percentage - Percentage.

Returns:

Validation dataset.

Split

public DatasetClassification Split(float percentage,
                                   java.lang.String name)

Split the percentange of the dataset in Trainning and the rest in Validation.

Parameters:

percentage - Percentage.

name - Name of the validation dataset.

Returns:

Validation dataset.

Shuffle

public void Shuffle()

Shuffle the dataset.

Shuffle

public void Shuffle(long seed)

Shuffle the dataset.

Parameters:

seed - Random seed.

Sort

public void Sort()

Sort the dataset in relation of the labels.

Standartize

public void Standartize()

Standartize all continuous data. x = (x - u) / s

getStatistics

public StatisticsDataset[] getStatistics()

Description copied from interface: IDataset

Get the statistics from the dataset.

Specified by:

getStatistics in interface IDataset<double[][],int[]>

Returns:

Statistics.

WriteAsCSV

public void WriteAsCSV(java.lang.String filename)

Write the dataset as CSV file.

Parameters:

filename - Filename.

WriteAsCSV

public void WriteAsCSV(java.lang.String filename,
                       int decimalPlaces)

Write the dataset as CSV file.

Parameters:

filename - Filename.

decimalPlaces - Decimal places for continous data.

WriteAsCSV

public void WriteAsCSV(java.lang.String filename,
                       int decimalPlaces,
                       char delimiter)

Write the dataset as CSV file.

Parameters:

filename - Filename.

decimalPlaces - Decimal places for continous data.

delimiter - Delimiter.

WriteAsCSV

public void WriteAsCSV(java.lang.String filename,
                       int decimalPlaces,
                       char delimiter,
                       java.lang.String newLine)

WriteAsCSV

public void WriteAsCSV(java.lang.String filename,
                       int decimalPlaces,
                       char delimiter,
                       java.lang.String newLine,
                       boolean writeHeader)

Write the dataset as CSV file.

Parameters:

filename - Filename.

decimalPlaces - Decimal places.

delimiter - Delimiter.

newLine - Newline char.

writeHeader - Save the file with no header information.

WriteAsARFF

public void WriteAsARFF(java.lang.String filename)

Write dataset as ARFF file.

Parameters:

filename - Filename.

WriteAsARFF

public void WriteAsARFF(java.lang.String filename,
                        int decimalPlaces)

Write dataset as ARFF file.

Parameters:

filename - Filename.

decimalPlaces - Decimal places.