net.returnvoid.analytics

Class ColorClustering

java.lang.Object

net.returnvoid.analytics.ColorClustering

All Implemented Interfaces:

Clustering

public class ColorClustering
extends java.lang.Object
implements Clustering

A class to store color clusterings. Instances of this class are meant to final. This means no changes of the clusters or to the content of any clusters should be made.

Author:

Diana Lange

Constructor Summary

Constructors

Constructor and Description
ColorClustering(ColorCluster[] clusters, int k) Builds a new clustering from a given set of clusters.

Method Summary

Modifier and Type	Method and Description
void	drawMeans(processing.core.PApplet parent, float x, float y, float w, float h) Draws the cluster means with horizontal striped within the given rectangular area.
void	drawMeans(processing.core.PGraphics g, float x, float y, float w, float h) Draws the cluster means with horizontal striped within the given rectangular area.
void	drawMembers(processing.core.PApplet parent, float x, float y, float w, float h) Draws the cluster members with horizontal striped within the given rectangular area.
void	drawMembers(processing.core.PGraphics g, float x, float y, float w, float h) Draws the cluster members with horizontal striped within the given rectangular area.
ColorCluster	get(int index) Returns the cluster with the given index.
ColorCluster	getCluster(int c) Returns the cluster which is the most similar to the given color.
ColorCluster	getCluster(java.lang.Object c) Returns the cluster which is the most similar to the given parameter.
ColorCluster	getCluster(RColor c) Returns the cluster which is the most similar to the given color.
int	getClusterIndex(int c) Returns the index of the cluster which is the most similar to the input color.
int	getClusterIndex(java.lang.Object c) Returns the index of the cluster which is the most similar to the input color.
int	getClusterIndex(RColor c) Returns the index of the cluster which is the most similar to the input color.
ColorCluster[]	getClusters() Returns all clusters of this clustering.
ColorDifferenceMeasure	getColorDifferenceMeasure() Returns the ColorDifferenceMeasure which is used to calculate the most similar cluster to a given color within the getCluster() and getClusterIndex() methods.
int	getK() The number of clusters to this clustering. k might differ from the getClusters().length, when empty clusters (clusters with no members) are automatically removed.
java.lang.Float	getLoss() Returns the loss of the clustering (= mean loss of all clusters).
java.lang.Float	getMaxLuminance() Returns the maximal luminance of this clustering.
RColor[]	getMeans() Returns all cluster means.
RColor[]	getMembers() Returns a set of colors which contains all members of all clusters of this clustering.
java.lang.Float	getMinLuminance() Returns the minimal luminance of this clustering.
int	size() The number of clusters for this clustering.
ColorPalette	toColorPalette() Builds a color palette from this clustering.
ColorPalette	toColorPalette(boolean useMeans) Builds a color palette from this clustering.

Methods inherited from class java.lang.Object

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

Constructor Detail

ColorClustering

public ColorClustering(ColorCluster[] clusters,
                       int k)

Builds a new clustering from a given set of clusters. The clusters should not be changed after the clustering has been build. All calculations / optimizations of the clusters should before building a clustering. The clusters will be sorted by luminance (first cluster = lowest luminance, last cluster = highest luminance).

Parameters:

clusters - The clusters with set means and cluster members. All clusters should uses the same ColorDifferenceMeasure.

k - The number of clusters of this clustering. k might differ from the clusters.length, when empty clusters (clusters with no members) are automatically removed.

Method Detail

getLoss

public java.lang.Float getLoss()

Description copied from interface: Clustering

Returns the loss of the clustering (= mean loss of all clusters).

Specified by:

getLoss in interface Clustering

Returns:

The loss of the clustering.

getK

public int getK()

Description copied from interface: Clustering

The number of clusters to this clustering. k might differ from the getClusters().length, when empty clusters (clusters with no members) are automatically removed.

Specified by:

getK in interface Clustering

Returns:

The number of clusters.

getMinLuminance

public java.lang.Float getMinLuminance()

Returns the minimal luminance of this clustering.

Returns:

The minimal luminance of this clustering [0, 100].

getMaxLuminance

public java.lang.Float getMaxLuminance()

Returns the maximal luminance of this clustering.

Returns:

The maximal luminance of this clustering [0, 100].

toColorPalette

public ColorPalette toColorPalette()

Builds a color palette from this clustering. The palette will consist of the the means of the clusters.

Returns:

A new (weighted) color palette.

toColorPalette

public ColorPalette toColorPalette(boolean useMeans)

Builds a color palette from this clustering. If useMeans is false: The palette will only contain colors of the cluster members (will choose a cluster member for each cluster that is the closest to the mean). Otherwise the means of the clusters will build the palette.

Parameters:

useMeans - A parameter which defines if the palette should consist of cluster means or a representative.

Returns:

A new (weighted) color palette.

getMeans

public RColor[] getMeans()

Returns all cluster means.

Returns:

The cluster means.

getMembers

public RColor[] getMembers()

Returns a set of colors which contains all members of all clusters of this clustering.

Returns:

The set of colors.

size

public int size()

The number of clusters for this clustering. This should be used to iterate over all clusters. Example:

 for (int i = 0; i < myClustering.size(); i++) {
        myClustering.get(i); // The cluster with index i
 }
 

Returns:

The number of clusters for this clustering.

get

public ColorCluster get(int index)

Returns the cluster with the given index. Make sure the index is within bounds, e.g. use the size() method to iterate over all clusters. Example:

 for (int i = 0; i < myClustering.size(); i++) {
        myClustering.get(i); // The cluster with index i
 }
 

Parameters:

index - The index of the cluster.

Returns:

The cluster with the given index.

getClusterIndex

public int getClusterIndex(java.lang.Object c)

Returns the index of the cluster which is the most similar to the input color.

Parameters:

c - A color.

Returns:

The index of the cluster which is the most similar to the input or -1 if the input wasn't a color.

getClusterIndex

public int getClusterIndex(int c)

Returns the index of the cluster which is the most similar to the input color.

Parameters:

c - A color (e.g 0xFFFF0000).

Returns:

The index of the cluster which is the most similar to the input or -1 if the input wasn't a color.

getClusterIndex

public int getClusterIndex(RColor c)

Returns the index of the cluster which is the most similar to the input color.

Parameters:

c - A color object.

Returns:

The index of the cluster which is the most similar to the input or -1 if the input wasn't a color.

getCluster

public ColorCluster getCluster(java.lang.Object c)

Description copied from interface: Clustering

Returns the cluster which is the most similar to the given parameter.

Specified by:

getCluster in interface Clustering

Parameters:

c - An element.

Returns:

The most similar cluster to the input or null if the input is not compatible to the data type of the cluster members.

getCluster

public ColorCluster getCluster(int c)

Returns the cluster which is the most similar to the given color.

Parameters:

c - A color (e.g 0xFFFF0000).

Returns:

The most similar cluster to the input.

getCluster

public ColorCluster getCluster(RColor c)

Returns the cluster which is the most similar to the given color.

Parameters:

c - A color.

Returns:

The most similar cluster to the input.

getColorDifferenceMeasure

public ColorDifferenceMeasure getColorDifferenceMeasure()

Returns the ColorDifferenceMeasure which is used to calculate the most similar cluster to a given color within the getCluster() and getClusterIndex() methods.

Returns:

The ColorDifferenceMeasure of this clustering.

getClusters

public ColorCluster[] getClusters()

Description copied from interface: Clustering

Returns all clusters of this clustering.

Specified by:

getClusters in interface Clustering

Returns:

The clusters.

drawMembers

public void drawMembers(processing.core.PGraphics g,
                        float x,
                        float y,
                        float w,
                        float h)

Draws the cluster members with horizontal striped within the given rectangular area.

Parameters:

g - A recorder.

x - The x-position (top left corner) of the rectangle.

y - The y-position (top left corner) of the rectangle.

w - The width of the rectangle.

h - The height of the rectangle.

drawMembers

public void drawMembers(processing.core.PApplet parent,
                        float x,
                        float y,
                        float w,
                        float h)

Draws the cluster members with horizontal striped within the given rectangular area.

Parameters:

parent - A recorder.

x - The x-position (top left corner) of the rectangle.

y - The y-position (top left corner) of the rectangle.

w - The width of the rectangle.

h - The height of the rectangle.

drawMeans

public void drawMeans(processing.core.PGraphics g,
                      float x,
                      float y,
                      float w,
                      float h)

Draws the cluster means with horizontal striped within the given rectangular area.

Parameters:

g - A recorder or null.

x - The x-position (top left corner) of the rectangle.

y - The y-position (top left corner) of the rectangle.

w - The width of the rectangle.

h - The height of the rectangle.

drawMeans

public void drawMeans(processing.core.PApplet parent,
                      float x,
                      float y,
                      float w,
                      float h)

Draws the cluster means with horizontal striped within the given rectangular area.

Parameters:

parent - A recorder or null.

x - The x-position (top left corner) of the rectangle.

y - The y-position (top left corner) of the rectangle.

w - The width of the rectangle.

h - The height of the rectangle.