 |
BUILT-IN MATHEMATICA SYMBOL
ClusteringComponents[array]
gives an array in which each element at the lowest level of array is replaced by an integer index representing the cluster in which the element lies.
ClusteringComponents[array, n]
finds at most n clusters.
ClusteringComponents[array, n, level]
finds clusters at the specified level in array.
ClusteringComponents[image]
finds clusters of pixels with similar values in image.
ClusteringComponents[image, n]
finds at most n clusters in image.
Details and Options
- ClusteringComponents[source, DistanceFunction->f] treats pairs of elements,
in source as being less similar when their distances 
are larger. - For images and real data, ClusteringComponents by default effectively uses the Euclidean distance function EuclideanDistance to determine the similarity of elements.
- For lists of True and False, ClusteringComponents by default uses a distance function based on the normalized fraction of elements that disagree.
- For lists of strings, ClusteringComponents by default uses a distance function based on the number of point changes needed to get from one string to another.
- Other distance functions can be specified by setting the DistanceFunction option. Possible settings are:
-
ManhattanDistance Manhattan or "city block" distance EuclideanDistance Euclidean distance SquaredEuclideanDistance squared Euclidean distance NormalizedSquaredEuclideanDistance normalized squared Euclidean distance CosineDistance angular cosine distance CorrelationDistance correlation coefficient distance - A Method option can be used to specify different methods of clustering. Possible settings include:
-
"Agglomerate" find clustering hierarchically "Optimize" find clustering by local optimization "KMeans" 
-means clustering algorithm"PAM" find clustering by partitioning around medoids - ClusteringComponents accepts a
option that is used to control the creation of the initial set of seeds. - ClusteringComponents also works with Image3D objects.
New in 8 | Last modified in 9
