stream (version 1.5-1)

DSC_Hierarchical: Hierarchical Micro-Cluster Reclusterer

Description

Macro Clusterer. Implementation of hierarchical clustering to recluster a set of micro-clusters.

Usage

DSC_Hierarchical(
  k = NULL,
  h = NULL,
  method = "complete",
  min_weight = NULL,
  description = NULL
)

Value

A list of class DSC, DSC_R, DSC_Macro, and DSC_Hierarchical. The list contains the following items:

description

The name of the algorithm in the DSC object.

RObj

The underlying R object.

Arguments

k

The number of desired clusters.

h

Height where to cut the dendrogram.

method

the agglomeration method to be used. This should be (an unambiguous abbreviation of) one of "ward", "single", "complete", "average", "mcquitty", "median" or "centroid".

min_weight

micro-clusters with a weight less than this will be ignored for reclustering.

description

optional character string to describe the clustering method.

Author

Michael Hahsler

Details

Please refer to hclust for more details on the behavior of the algorithm.

Note that this clustering cannot be updated iteratively and every time it is used for (re)clustering, the old clustering is deleted.

See Also

DSC, DSC_Macro

Examples

# Cassini dataset
stream <- DSD_mlbenchGenerator("cassini")

# Use hierarchical clustering to recluster micro-clusters
dbstream <- DSC_DBSTREAM(r = .05)
update(dbstream, stream, 500)

# reclustering using single-link and specifying k
hc <- DSC_Hierarchical(k = 3, method = "single")
recluster(hc, dbstream)
hc
plot(hc, stream, type = "both")

# reclustering by specifying height
hc <- DSC_Hierarchical(h = .2, method = "single")
recluster(hc, dbstream)
hc
plot(hc, stream, type = "both")

# For comparison we use hierarchical clustering directly on the data
# Note: hierarchical clustering is not a data stream clustering algorithm!
hc <- DSC_Hierarchical(k = 3, method = "single")
update(hc, stream, 500)
plot(hc, stream)