de.jstacs.clustering.hierachical

Class ClusterTree<T>

java.lang.Object

de.jstacs.clustering.hierachical.ClusterTree<T>

Type Parameters:

T - the type of the leaves

All Implemented Interfaces:

Storable

public class ClusterTree<T>
extends Object
implements Storable

Class for a generic cluster tree with leaves of type T. Cluster trees for a given set of leaf elements may be obtained using Hclust.

Author:

Jan Grau

Constructor Summary

Constructors

Constructor and Description
ClusterTree(double distance, int originalIndex, ClusterTree<T>... subTrees) Creates a new cluster tree with supplied sub-trees and given distance.
ClusterTree(StringBuffer xml) Creates a cluster tree from its XML representation
ClusterTree(T leaf, int originalIndex) Creates a new cluster tree for a given leaf element (i.e., the tree comprises just this leaf) with the supplied index in the set of cluster elements

Method Summary

Modifier and Type	Method and Description
<S> ClusterTree<S>	dropBelow(IntList rootOriginalIndexes, S[] newElements) Removes all sub-trees below the inner nodes identified by the original indexes supplied and creates new leaf nodes instead, which obtain the supplied leaf elements.
T[]	getClusterElements() Returns the elements at all leaves in this cluster tree, in the order of the leaves, from left to right.
double	getDistance() Returns the distance between the child trees of this cluster tree root node.
ClusterTree<Integer>	getIndexTree() Returns a cluster tree with identical structure as this cluster tree but with all leaves replaced by integer leaves holding the corresponding original indices.
ClusterTree<T>[]	getLeaves() Returns all leaves of this cluster tree as ClusterTree objects comprising just the corresponding leaf element
double	getMaximumDistance() Returns the maximum distance of trees under this root node.
double	getMinimumDistance() Returns the minimum distance of trees under this root node.
int	getMinimumOriginalIndex() Returns the minimum original index in this cluster tree.
int	getNumberOfElements() Returns the number of leaves in this cluster tree.
int	getOriginalIndex() Returns the original index of the root node of this cluster tree
ClusterTree<T>[]	getSubTrees() Returns the sub-trees of this cluster tree root node
boolean	isLeaf() Returns true if this cluster tree comprises just a leaf.
void	leafOrder(double[][] dmat) Orders the leaves of this cluster tree such that adjacent nodes have minimal distance.
void	reverseOrder() Reverses the order of the child trees of this cluster tree root node.
void	setElements() Sets the copy references of the leave nodes of this cluster tree to the elements of its leaves in the current order.
void	setOriginalIndex(int used) Sets the original index (e.g., if elements have been removed from the tree) referring to indexes in the distance matrix that has been used to build a tree.
String	toNewick() Returns a string representation of this cluster tree in newick format.
String	toString()
StringBuffer	toXML() This method returns an XML representation as StringBuffer of an instance of the implementing class.

Methods inherited from class java.lang.Object

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

Constructor Detail

ClusterTree

public ClusterTree(T leaf,
                   int originalIndex)

Creates a new cluster tree for a given leaf element (i.e., the tree comprises just this leaf) with the supplied index in the set of cluster elements

Parameters:

leaf - the leaf element

originalIndex - the index of the leaf element in the complete set of cluster elements

ClusterTree

public ClusterTree(double distance,
                   int originalIndex,
                   ClusterTree<T>... subTrees)

Creates a new cluster tree with supplied sub-trees and given distance. The original index may be a virtual index that helps to identify this inner (or root) node of the cluster tree later on.

Parameters:

distance - the distance between the sub-trees

originalIndex - the original index of this inner (or root) node

subTrees - the sub-trees

ClusterTree

public ClusterTree(StringBuffer xml)
            throws NonParsableException

Creates a cluster tree from its XML representation

Parameters:

xml - the XML representation

Throws:

NonParsableException - if XML could not be parsed

Method Detail

getIndexTree

public ClusterTree<Integer> getIndexTree()

Returns a cluster tree with identical structure as this cluster tree but with all leaves replaced by integer leaves holding the corresponding original indices.

Returns:

the index cluster tree

See Also:

ClusterTree(Object, int)

leafOrder

public void leafOrder(double[][] dmat)

Orders the leaves of this cluster tree such that adjacent nodes have minimal distance. Implements the algorithm of Ziv Bar-Joseph et al. "K-ary Clustering with Optimal Leaf Ordering for Gene Expression Data"

Parameters:

dmat - the distance matrix that has also been used to build the tree (using the same indexes)

getOriginalIndex

public int getOriginalIndex()

Returns the original index of the root node of this cluster tree

Returns:

the original index

reverseOrder

public void reverseOrder()

Reverses the order of the child trees of this cluster tree root node.

toXML

public StringBuffer toXML()

Description copied from interface: Storable

This method returns an XML representation as StringBuffer of an instance of the implementing class.

Specified by:

toXML in interface Storable

Returns:

the XML representation

setElements

public void setElements()

Sets the copy references of the leave nodes of this cluster tree to the elements of its leaves in the current order.

getSubTrees

public ClusterTree<T>[] getSubTrees()

Returns the sub-trees of this cluster tree root node

Returns:

the sub-trees

getDistance

public double getDistance()

Returns the distance between the child trees of this cluster tree root node.

Returns:

the distance

getMaximumDistance

public double getMaximumDistance()

Returns the maximum distance of trees under this root node. In this implementation identical to getDistance(). Basically used for plotting the tree structure.

Returns:

the distance

getMinimumDistance

public double getMinimumDistance()

Returns the minimum distance of trees under this root node. Basically used for plotting the tree structure.

Returns:

the distance

getClusterElements

public T[] getClusterElements()

Returns the elements at all leaves in this cluster tree, in the order of the leaves, from left to right.

Returns:

the elements

getNumberOfElements

public int getNumberOfElements()

Returns the number of leaves in this cluster tree.

Returns:

the number of leaves

toString

public String toString()

Overrides:

toString in class Object

getMinimumOriginalIndex

public int getMinimumOriginalIndex()

Returns the minimum original index in this cluster tree.

Returns:

the minimum original index

isLeaf

public boolean isLeaf()

Returns true if this cluster tree comprises just a leaf.

Returns:

if this tree is a leaf

getLeaves

public ClusterTree<T>[] getLeaves()

Returns all leaves of this cluster tree as ClusterTree objects comprising just the corresponding leaf element

Returns:

the leaf trees

dropBelow

public <S> ClusterTree<S> dropBelow(IntList rootOriginalIndexes,
                                    S[] newElements)

Removes all sub-trees below the inner nodes identified by the original indexes supplied and creates new leaf nodes instead, which obtain the supplied leaf elements. The modified tree is returned.

Type Parameters:

S - the type of the new leaves

Parameters:

rootOriginalIndexes - the original indexes of the inner nodes to be replaced

newElements - the leaf element replacements

Returns:

the modified tree

toNewick

public String toNewick()

Returns a string representation of this cluster tree in newick format.

Returns:

the string representation

setOriginalIndex

public void setOriginalIndex(int used)

Sets the original index (e.g., if elements have been removed from the tree) referring to indexes in the distance matrix that has been used to build a tree.

Parameters:

used - the new index