Package ai.libs.jaicore.ml.tsc.util

Class WekaUtil

java.lang.Object

ai.libs.jaicore.ml.tsc.util.WekaUtil

public class WekaUtil
extends java.lang.Object

WekaUtil

Method Summary

Modifier and Type	Method	Description
static void	buildWekaClassifierFromSimplifiedTS(weka.classifiers.Classifier classifier, TimeSeriesDataset timeSeriesDataset)	Trains a given Weka classifier using the simplified time series data set timeSeriesDataset.
static <L> void	buildWekaClassifierFromTS(weka.classifiers.Classifier classifier, TimeSeriesDataset<L> timeSeriesDataset)	Trains a given Weka classifier using the time series data set timeSeriesDataset.
static weka.core.Instances	indArrayToWekaInstances(org.nd4j.linalg.api.ndarray.INDArray matrix)	Converts an INDArray matrix (number of instances x number of attributes) to Weka instances without any class attribute.
static weka.core.Instances	matrixToWekaInstances(double[][] matrix)	Converts a double[][] matrix (number of instances x number of attributes) to Weka instances without any class attribute.
static weka.core.Instances	simplifiedTimeSeriesDatasetToWekaInstances(TimeSeriesDataset dataSet)	Converts a given simplified TimeSeriesDataset object to a Weka Instances object.
static weka.core.Instances	simplifiedTimeSeriesDatasetToWekaInstances(TimeSeriesDataset dataSet, java.util.List<java.lang.String> classValues)	Converts a given simplified TimeSeriesDataset object to a Weka Instances object.
static weka.core.Instance	simplifiedTSInstanceToWekaInstance(double[] instance)	Maps an univariate simplified time series instance to a Weka instance.
static <L> weka.core.Instances	timeSeriesDatasetToWekaInstances(TimeSeriesDataset<L> dataSet)	Converts a given TimeSeriesDataset object to a Weka Instances object.
static weka.core.Instance	tsInstanceToWekaInstance(TimeSeriesInstance<?> instance)	Maps a time series instance to a Weka instance.
static org.nd4j.linalg.api.ndarray.INDArray	wekaInstancesToINDArray(weka.core.Instances instances, boolean keepClass)	Converts Weka instances to an INDArray matrix.

Methods inherited from class java.lang.Object

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

Method Detail

tsInstanceToWekaInstance

public static weka.core.Instance tsInstanceToWekaInstance(TimeSeriesInstance<?> instance)

Maps a time series instance to a Weka instance.

Parameters:

instance - The time series instance storing the time series data and the target value

Returns:

Returns the Weka instance containing the time series data and the class information.

simplifiedTSInstanceToWekaInstance

public static weka.core.Instance simplifiedTSInstanceToWekaInstance(double[] instance)

Maps an univariate simplified time series instance to a Weka instance.

Parameters:

instance - The time series instance storing the time series data

Returns:

Returns the Weka instance containing the time series

buildWekaClassifierFromTS

public static <L> void buildWekaClassifierFromTS(weka.classifiers.Classifier classifier,
                                                 TimeSeriesDataset<L> timeSeriesDataset)
                                          throws TrainingException

Trains a given Weka classifier using the time series data set timeSeriesDataset.

Parameters:

classifier - The Weka weka.Classifier object

timeSeriesDataset - The time series data set which is transformed to Weka instances used for the training

Throws:

TrainingException - Throws exception if the training could not be finished successfully

buildWekaClassifierFromSimplifiedTS

public static void buildWekaClassifierFromSimplifiedTS(weka.classifiers.Classifier classifier,
                                                       TimeSeriesDataset timeSeriesDataset)
                                                throws TrainingException

Trains a given Weka classifier using the simplified time series data set timeSeriesDataset.

Parameters:

classifier - The Weka weka.Classifier object

timeSeriesDataset - The time series data set which is transformed to Weka instances used for the training

Throws:

TrainingException - Throws exception if the training could not be finished successfully

wekaInstancesToINDArray

public static org.nd4j.linalg.api.ndarray.INDArray wekaInstancesToINDArray(weka.core.Instances instances,
                                                                           boolean keepClass)

Converts Weka instances to an INDArray matrix.

Parameters:

instances - Weka instances to be converted.

keepClass - Determines whether the class attribute should be stored in the result matrix

Returns:

Returns an INDArray consisting of all instances with the shape (number instances x number attributes)

timeSeriesDatasetToWekaInstances

public static <L> weka.core.Instances timeSeriesDatasetToWekaInstances(TimeSeriesDataset<L> dataSet)

Converts a given TimeSeriesDataset object to a Weka Instances object. Works with CategoricalAttributeType target values.

Parameters:

dataSet - Data set which is transformed

Returns:

Transformed Weka Instances object

simplifiedTimeSeriesDatasetToWekaInstances

public static weka.core.Instances simplifiedTimeSeriesDatasetToWekaInstances(TimeSeriesDataset dataSet)

Converts a given simplified TimeSeriesDataset object to a Weka Instances object.

Parameters:

dataSet - Data set which is transformed

Returns:

Transformed Weka Instances object

simplifiedTimeSeriesDatasetToWekaInstances

public static weka.core.Instances simplifiedTimeSeriesDatasetToWekaInstances(TimeSeriesDataset dataSet,
                                                                             java.util.List<java.lang.String> classValues)

Converts a given simplified TimeSeriesDataset object to a Weka Instances object.

Parameters:

dataSet - Data set which is transformed

Returns:

Transformed Weka Instances object

indArrayToWekaInstances

public static weka.core.Instances indArrayToWekaInstances(org.nd4j.linalg.api.ndarray.INDArray matrix)

Converts an INDArray matrix (number of instances x number of attributes) to Weka instances without any class attribute.

Parameters:

matrix - INDArray matrix storing all the attribute values of the instances

Returns:

Returns the Weka Instances object consisting of all instances and the attribute values

matrixToWekaInstances

public static weka.core.Instances matrixToWekaInstances(double[][] matrix)

Converts a double[][] matrix (number of instances x number of attributes) to Weka instances without any class attribute.

Parameters:

matrix - The double[][] matrix storing all the attribute values of the instances

Returns:

Returns the Weka Instances object consisting of all instances and the attribute values