Package | Description |
---|---|
at.tuwien.ifs.somtoolbox.apps.trainer | |
at.tuwien.ifs.somtoolbox.clustering.functions | |
at.tuwien.ifs.somtoolbox.data |
Classes in this package implement reading, storing and providing of different data needed for the SOM, e.g.
|
at.tuwien.ifs.somtoolbox.data.distance | |
at.tuwien.ifs.somtoolbox.layers |
Provides the basic classes constituting SOM-based neural networks.
|
at.tuwien.ifs.somtoolbox.layers.metrics |
This package contains the metrics used for distance calculation during training and for mapping data onto maps.
|
at.tuwien.ifs.somtoolbox.layers.quality |
Classes in this package implement various quality measures, indicating the quality of the SOM mapping.
|
at.tuwien.ifs.somtoolbox.visualization |
Provides classes creating visualisations of trained SOMs.
|
at.tuwien.ifs.somtoolbox.visualization.clustering |
Contains classes implementing clustering methods on the SOM.
|
Modifier and Type | Field and Description |
---|---|
private ClassComboBoxModel<DistanceMetric> |
SOMTrainer.cmbMetricModel |
Modifier and Type | Method and Description |
---|---|
private ClassComboBoxModel<DistanceMetric> |
SOMTrainer.getCmbMetricModel() |
Modifier and Type | Field and Description |
---|---|
protected DistanceMetric |
UnitClusteringFunction.metric |
protected DistanceMetric |
GeneralUnitPNodeClusteringFunction.metric |
protected DistanceMetric |
DoubleVector2DDistance.metric |
Constructor and Description |
---|
DoubleVector2DDistance(DistanceMetric metric) |
GeneralUnitPNodeClusteringFunction(DistanceMetric metric) |
UnitClusteringFunction(DistanceMetric metric) |
Modifier and Type | Method and Description |
---|---|
InputDatum[] |
AbstractSOMLibSparseInputData.getByNameDistanceSorted(double[] vector,
Collection<String> inputNames,
DistanceMetric metric)
Retrieves the
InputDatum corresponding to the given input names, and sorted by their distance to the
given vector. |
ArrayList<InputDistance> |
AbstractSOMLibSparseInputData.getDistances(int inputIndex,
DistanceMetric metric)
Returns the distances to the index of the given vector of the dataset.
|
SmallestElementSet<InputDistance> |
AbstractSOMLibSparseInputData.getNearestDistances(int inputIndex,
int neighbours,
DistanceMetric metric) |
InputDatum[] |
AbstractSOMLibSparseInputData.getNearestN(double[] vector,
DistanceMetric metric,
int number)
Retrieves the given number of
InputDatum that are closest to the given vector. |
InputDatum[] |
AbstractSOMLibSparseInputData.getNearestN(int inputIndex,
DistanceMetric metric,
int number)
Returns the n nearest input vectors for the index of the given vector of the dataset.
|
InputDatum[] |
AbstractSOMLibSparseInputData.getNearestNUnsorted(int inputIndex,
DistanceMetric metric,
int number) |
void |
AbstractSOMLibSparseInputData.initDistanceMatrix(DistanceMetric metric)
Calculates the
AbstractSOMLibSparseInputData.distanceMatrix - careful, this is a lengthy process and should be done only if needed. |
double |
SimpleMatrixInputData.mqe0(DistanceMetric metric) |
double |
SOMLibSparseInputData.mqe0(DistanceMetric metric) |
double |
RandomAccessFileSOMLibInputData.mqe0(DistanceMetric metric) |
double |
InputData.mqe0(DistanceMetric metric)
Calculates the mean quantisation error of the top-level unit.
|
double |
DataBaseSOMLibSparseInputData.mqe0(DistanceMetric metric) |
void |
AbstractSOMLibSparseInputData.transformValues(DistanceMetric metric)
Calculates the matrix of
AbstractSOMLibSparseInputData.transformedVectors using transformVector(double[]) of
the given metric. |
Modifier and Type | Field and Description |
---|---|
protected DistanceMetric |
InputVectorDistanceMatrix.metric |
Modifier and Type | Method and Description |
---|---|
DistanceMetric |
RandomAccessFileInputVectorDistanceMatrix.getMetric() |
DistanceMetric |
InputVectorDistanceMatrix.getMetric() |
Modifier and Type | Method and Description |
---|---|
private static PrintWriter |
DistanceMatrixWriter.printSOMLibHeader(int numVectors,
String fileName,
DistanceMetric metric,
boolean gzip) |
static void |
DistanceMatrixWriter.writeOrangeFileInputVectorDistanceMatrix(InputData data,
String fileName,
DistanceMetric metric)
Write input distance matrix to an ASCII file for the Orange data mining toolkit ((http://www.ailab.si/orange/),
computing distances on the fly.
|
static void |
DistanceMatrixWriter.writePlainFileInputVectorDistanceMatrix(InputData data,
String fileName,
DistanceMetric metric)
Write input distance matrix to an ASCII file in plain format, computing distances on the fly.
|
static void |
DistanceMatrixWriter.writeRandomAccessFileInputVectorDistanceMatrix(double[][] distances,
String fileName,
DistanceMetric metric)
Write pre-calculated input distance matrix to a binary file.
|
static void |
DistanceMatrixWriter.writeRandomAccessFileInputVectorDistanceMatrix(InputData data,
String fileName,
DistanceMetric metric)
Write input distance matrix to a binary file, computing distances on the fly.
|
static void |
DistanceMatrixWriter.writeSOMLibFileInputVectorDistanceMatrix(double[][] distances,
String fileName,
DistanceMetric metric,
boolean gzip)
Write pre-calculated input distance matrix to an ASCII file in SOMLib format.
|
static void |
DistanceMatrixWriter.writeSOMLibFileInputVectorDistanceMatrix(InputData data,
String fileName,
DistanceMetric metric)
Write input distance matrix to ASCII file, computing distances on the fly.
|
static void |
DistanceMatrixWriter.writeSOMLibFileInputVectorDistanceMatrix(InputData data,
String fileName,
DistanceMetric metric,
boolean gzip)
Write input distance matrix to ASCII file, computing distances on the fly.
|
Constructor and Description |
---|
AbstractMemoryInputVectorDistanceMatrix(InputData data,
DistanceMetric metric)
Constructs the distance matrix by computing the distances on the fly.
|
FullMemoryInputVectorDistanceMatrix(InputData data,
DistanceMetric metric) |
LeightWeightMemoryInputVectorDistanceMatrix(InputData data,
DistanceMetric metric) |
Modifier and Type | Field and Description |
---|---|
protected DistanceMetric |
GrowingLayer.metric |
private DistanceMetric |
GrowingCellLayer.metric
Distance Metric used for GrowingCellStructures
|
Modifier and Type | Method and Description |
---|---|
DistanceMetric |
Layer.getMetric()
Returns the metric used for distance calculation.
|
DistanceMetric |
GrowingLayer.getMetric()
Calculates and returns the mean quantization error of the map based on the quantization errors of the single
units.
|
DistanceMetric |
GrowingCellLayer.getMetric() |
Modifier and Type | Method and Description |
---|---|
Unit |
GrowingLayer.getWinner(InputDatum input,
DistanceMetric metric)
Returns the winner unit for a given input datum specified by argument
input . |
Unit[] |
GrowingLayer.getWinners(InputDatum input,
int num,
DistanceMetric metric)
Returns a number of best-matching units sorted by distance (ascending) for a given input datum.
|
void |
AdaptiveCoordinatesVirtualLayer.updateUnitsVirtualSpacePos(Unit[][][] units,
DistanceMetric metric,
Unit winner,
InputDatum input,
int curIteration)
Updates the virtual space position of all map units with respect to the input datum and the according winner
unit.
|
Modifier and Type | Class and Description |
---|---|
class |
AbstractMetric
Implements a static method for metric instantiation and a method for mean vector calculation.
|
class |
AbstractWeightedMetric |
class |
CosineMetric
Implements the cosine metric, defined for two vectors d1 and d2 as d1xd2 / (|d1|*|d2|).
|
class |
L1Metric
Implements the L1 or city block metric.
|
class |
L2Metric
Implements the L2 or Euclidean metric.
|
class |
L2MetricFast
Implements a fast version of the L2 or Euclidean metric, by not taking the square root.
|
class |
L2MetricSparse
Implements the L2 or Euclidean metric, considering only those values for the distance calculation that have non-zero
values for the first, second or both vectors, depending on the initialisation mode.
|
class |
L2MetricWeighted |
class |
LInfinityMetric
Implements the L-Infinity metric,defined for two vectors x and y as max( |xi-yi| ), i = 1,...,|x|.
|
class |
LnAlphaMetric |
class |
LnMetric
Generic Ln metric.
|
class |
MahalanobisMetric
Implements the Mahalanobis distance metric.
|
class |
MissingValueMetricWrapper
A wrapper class around other distance metrics, modifying the distance computation in such a way that only vector
attributes that are not missing (indicated by
InputData.MISSING_VALUE are considered.When instantiating using the empty constructor MissingValueMetricWrapper.MissingValueMetricWrapper() the default metric
MissingValueMetricWrapper.DEFAULT_METRIC is used. |
class |
MnemonicSOMMetric
A metric for mnemonic SOMs.
|
Modifier and Type | Field and Description |
---|---|
private DistanceMetric |
MissingValueMetricWrapper.metric |
private static DistanceMetric[] |
Metrics.singleton |
Modifier and Type | Method and Description |
---|---|
static DistanceMetric[] |
Metrics.getAvailableMetrics() |
static DistanceMetric |
AbstractMetric.instantiate(String mName)
Instantiates a certain distance metric class specified by argument
mName .Note: for backwards compatibility, if the metric name contains the package prefix at.ec3.somtoolbox ,
this will be replaced by at.tuwien.ifs.somtoolbox . |
static DistanceMetric |
AbstractMetric.instantiateNice(String metricName)
Same as
AbstractMetric.instantiate(String) , but tries to get the metric with the specified name, and then with the
package prefix, and throwing only a SOMToolboxException with the root cause nested. |
Modifier and Type | Method and Description |
---|---|
int |
AbstractMetric.compareTo(DistanceMetric o) |
protected static void |
AbstractMetric.performanceTest(DistanceMetric metric,
int dim)
Can be used to do some performance testing to compare colt vs.
|
void |
MissingValueMetricWrapper.setMetric(DistanceMetric metric) |
Constructor and Description |
---|
MissingValueMetricWrapper(DistanceMetric metric) |
Modifier and Type | Field and Description |
---|---|
(package private) DistanceMetric |
TopographicProduct.metric |
(package private) DistanceMetric |
Trustworthiness_NeighborhoodPreservation.metric |
Modifier and Type | Field and Description |
---|---|
private DistanceMetric |
NeighbourhoodGraph.metric |
Modifier and Type | Field and Description |
---|---|
private DistanceMetric |
Cluster.distanceFunction |
Modifier and Type | Method and Description |
---|---|
private void |
KMeans.initClustersEqualNumbers(DistanceMetric distanceFunction)
cluster centres are initialised by equally sized random chunks of the input data when there's 150 instances, we
assign 50 chosen randomly to each cluster and calculate its centre from these (the last cluster might be larger
if numInstances mod k < 0)
|
private void |
KMeans.initClustersLinearly(DistanceMetric distanceFunction)
This one does linear initialisation.
|
private void |
KMeans.initClustersLinearlyOnInstances(DistanceMetric distanceFunction)
like
KMeans.initClustersLinearly(DistanceMetric) , but after computing the exact linear point, rather finds &
uses the closest instance from the data set as centroid. |
private void |
KMeans.initClustersRandomly(DistanceMetric distanceFunction)
Calculate random centroids for each cluster.
|
private void |
KMeans.initClustersRandomlyOnInstances(DistanceMetric distanceFunction)
Take random points from the input data as centroids.
|
Constructor and Description |
---|
Cluster(DistanceMetric distanceFunction) |
Cluster(double[] centroid,
DistanceMetric distanceFunction) |
KMeans(int k,
double[][] data,
KMeans.InitType initialisation,
DistanceMetric distanceFunction)
Construct a new K-Means bugger.
|