Package elki.data
Basic classes for different data types, database object types and label types.

Interface Summary Interface Description FeatureVector<D> Generic FeatureVector class that can contain any type of data (i.e. numerical or categorical attributes).FeatureVector.Factory<V extends FeatureVector<? extends D>,D> Factory API for this feature vector.NumberVector Interface NumberVector defines the methods that should be implemented by any Object that is element of a real vector space of type N.NumberVector.Factory<V extends NumberVector> Factory API for this feature vector.SparseFeatureVector<D> Extended interface for sparse feature vector types.SparseNumberVector Combines the SparseFeatureVector and NumberVector.SparseNumberVector.Factory<V extends SparseNumberVector> Factory for sparse number vectors: make from a dimvalue map. 
Class Summary Class Description Bit A boolean number type.BitVector Vector using a dense bit set encoding, based onlong[]
storage.BitVector.Factory Factory for bit vectors.BitVector.Factory.Par Parameterization class.BitVector.ShortSerializer Serialization class for dense integer vectors with up toShort.MAX_VALUE
dimensions, by using a short for storing the dimensionality.ByteVector Vector usingbyte[]
storage.ByteVector.Factory Factory for Byte vectors.ByteVector.Factory.Par Parameterization class.ByteVector.ShortSerializer Serialization class for dense Byte vectors with up toShort.MAX_VALUE
dimensions, by using a short for storing the dimensionality.ByteVector.SmallSerializer Serialization class for dense Byte vectors with up to 127 dimensions, by using a byte for storing the dimensionality.ClassLabel A ClassLabel to identify a certain class of objects that is to discern from other classes by a classifier.ClassLabel.Factory<L extends ClassLabel> Class label factory.Cluster<M extends Model> Generic cluster class, that may or not have hierarchical information.Clustering<M extends Model> Result class for clusterings.DoubleVector Vector type usingdouble[]
storage for real numbers.DoubleVector.Factory Factory for Double vectors.DoubleVector.Factory.Par Parameterization class.DoubleVector.ShortSerializer Serialization class for dense double vectors with up toShort.MAX_VALUE
dimensions, by using a short for storing the dimensionality.DoubleVector.SmallSerializer Serialization class for dense double vectors with up to 127 dimensions, by using a byte for storing the dimensionality.DoubleVector.VariableSerializer Serialization class for variable dimensionality by using VarInt encoding.ExternalID External ID objects.FloatVector Vector type usingfloat[]
storage, thus needing approximately half as much memory asDoubleVector
.FloatVector.Factory Factory for float vectors.FloatVector.Factory.Par Parameterization class.FloatVector.ShortSerializer Serialization class for dense float vectors with up toShort.MAX_VALUE
dimensions, by using a short for storing the dimensionality.FloatVector.SmallSerializer Serialization class for dense float vectors with up to 127 dimensions, by using a byte for storing the dimensionality.FloatVector.VariableSerializer Serialization class for variable dimensionality by using VarInt encoding.HierarchicalClassLabel A HierarchicalClassLabel is a ClassLabel to reflect a hierarchical structure of classes.HierarchicalClassLabel.Factory Factory class.HyperBoundingBox HyperBoundingBox represents a hyperrectangle in the multidimensional space.IntegerVector Vector type usingint[]
storage.IntegerVector.Factory Factory for integer vectors.IntegerVector.Factory.Par Parameterization class.IntegerVector.ShortSerializer Serialization class for dense integer vectors with up toShort.MAX_VALUE
dimensions, by using a short for storing the dimensionality.IntegerVector.SmallSerializer Serialization class for dense integer vectors with up to 127 dimensions, by using a byte for storing the dimensionality.IntegerVector.VariableSerializer Serialization class for variable dimensionality by using VarInt encoding.LabelList A list of string labels.LabelList.Serializer Serialization class.ModifiableHyperBoundingBox MBR class allowing modifications (as opposed toHyperBoundingBox
).OneDimensionalDoubleVector Specialized class implementing a onedimensional double vector without using an array.OneDimensionalDoubleVector.Factory Factory class.OneDimensionalDoubleVector.Factory.Par Parameterization class.ShortVector Vector type usingshort[]
storage.ShortVector.Factory Factory for Short vectors.ShortVector.Factory.Par Parameterization class.ShortVector.ShortSerializer Serialization class for dense Short vectors with up toShort.MAX_VALUE
dimensions, by using a short for storing the dimensionality.ShortVector.VariableSerializer Serialization class for variable dimensionality by using VarInt encoding.SimpleClassLabel A simple class label casting a String as it is as label.SimpleClassLabel.Factory Factory class.SimpleClassLabel.Serializer Serialization class.SparseByteVector Sparse vector type, usingbyte[]
for storing the values, andint[]
for storing the indexes, approximately 5 bytes per nonzero value (limited to 128..+127).SparseByteVector.Factory Factory class.SparseByteVector.Factory.Par Parameterization class.SparseByteVector.VariableSerializer Serialization class using VarInt encodings.SparseDoubleVector Sparse vector type, usingdouble[]
for storing the values, andint[]
for storing the indexes, approximately 12 bytes per nonzero value.SparseDoubleVector.Factory Factory class.SparseDoubleVector.Factory.Par Parameterization class.SparseDoubleVector.VariableSerializer Serialization class using VarInt encodings.SparseFloatVector Sparse vector type, usingfloat[]
for storing the values, andint[]
for storing the indexes, approximately 8 bytes per nonzero value.SparseFloatVector.Factory Factory class.SparseFloatVector.Factory.Par Parameterization class.SparseFloatVector.VariableSerializer Serialization class using VarInt encodings.SparseIntegerVector Sparse vector type, usingint[]
for storing the values, andint[]
for storing the indexes, approximately 8 bytes per nonzero integer value.SparseIntegerVector.Factory Factory class.SparseIntegerVector.Factory.Par Parameterization class.SparseIntegerVector.VariableSerializer Serialization class using VarInt encodings.SparseShortVector Sparse vector type, usingshort[]
for storing the values, andint[]
for storing the indexes, approximately 6 bytes per nonzero value.SparseShortVector.Factory Factory class.SparseShortVector.Factory.Par Parameterization class.SparseShortVector.VariableSerializer Serialization class using VarInt encodings.Subspace Represents a subspace of the original data space.VectorUtil Utility functions for use with vectors.VectorUtil.SortDBIDsBySingleDimension Compare number vectors by a single dimension.VectorUtil.SortVectorsBySingleDimension Compare number vectors by a single dimension.