[ VIGRA Homepage | Function Index | Class Index | Namespaces | File List | Main Page ]

details ClusteringOptions Class Reference VIGRA

Options object for hierarchical clustering. More...

#include <vigra/hierarchical_clustering.hxx>

Public Member Functions

ClusteringOptionsmaxMergeDistance (double val)
 
ClusteringOptionsminRegionCount (size_t count)
 
ClusteringOptionsnodeFeatureImportance (double val)
 
ClusteringOptionsnodeFeatureMetric (metrics::MetricType metric)
 
ClusteringOptionssizeImportance (double val)
 
ClusteringOptionsverbose (bool val=true)
 

Detailed Description

Options object for hierarchical clustering.

#include <vigra/hierarchical_clustering.hxx>
Namespace: vigra

This class allows to set various parameters of hierarchicalClustering(). See there for usage examples.

Examples:
graph_agglomerative_clustering.cxx.

Member Function Documentation

ClusteringOptions& minRegionCount ( size_t  count)

Stop merging when the number of clusters reaches this threshold.

Default: 1 (don't stop early)

ClusteringOptions& maxMergeDistance ( double  val)

Stop merging when the weight of the cheapest edge exceeds this threshold.

Default: infinity (don't stop early)

ClusteringOptions& nodeFeatureImportance ( double  val)

Importance of node features relative to edge weights.

Must be between 0 and 1, with 0 meaning that node features are ignored, and 1 meaning that edge weights are ignored.

Default: 0.5 (equal importance)

ClusteringOptions& sizeImportance ( double  val)

Importance of node size.

Must be between 0 and 1, with 0 meaning that size is ignored, and 1 meaning that the algorithm prefers to keep cluster sizes balanced.

Default: 1.0 (prefer like-sized clusters)

ClusteringOptions& nodeFeatureMetric ( metrics::MetricType  metric)

Metric to be used when transforming node features into cluster distances.

The cluster (= node) distance is the respective norm of the difference vector between the corresponding node feature vectors.

Default: metrics::ManhattanMetric (L1-norm of the feature difference)

ClusteringOptions& verbose ( bool  val = true)

Display progress information.

Default: false


The documentation for this class was generated from the following file:

© Ullrich Köthe (ullrich.koethe@iwr.uni-heidelberg.de)
Heidelberg Collaboratory for Image Processing, University of Heidelberg, Germany

html generated using doxygen and Python
vigra 1.11.1 (Fri May 19 2017)