stream (version 1.5-1)

SampleDSO-class: Update a Data Stream Clustering Model

Description

Update a clustering model by clustering a number of input points from a data stream into a clustering object.

Value

The updated model is returned invisibly for reassignment (however, this is not necessary).

To obtain the updated model for a DSC (data stream clustering model), call get_centers() on the DSC object.

Arguments

object

an object of a subclass of DST (data stream mining task).

dsd

a DSD object (data stream).

n

number of points to cluster.

verbose

report progress.

block

maximal number of data points passed on at once to the algorithm. This only is used since R loops are very slow.

...

extra arguments for clusterer.

Author

Michael Hahsler

Details

update takes n times a single data points out of the DSD updates the model in object. Note that update directly modifies the object (which is a reference class) and thus the result does not need to be reassigned to the object name.

%cluster is the low level implementation of updating a %data stream clustering model and is called by update.

See Also

DSC, DSD, and animation for producing an animation of the clustering process.

Examples

stream <- DSD_Gaussians(k=3)
dstream <- DSC_DStream(gridsize=.05)  

update(dstream, stream, 500)
plot(dstream, stream)