1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
|
\name{aggExCluster}
\docType{methods}
\alias{aggExCluster}
\alias{aggexcluster}
\alias{aggExCluster-methods}
\alias{aggExCluster,matrix,missing-method}
\alias{aggExCluster,matrix,ExClust-method}
\alias{aggExCluster,Matrix,missing-method}
\alias{aggExCluster,Matrix,ExClust-method}
\alias{aggExCluster,missing,ExClust-method}
\alias{aggExCluster,function,ANY-method}
\alias{aggExCluster,character,ANY-method}
\title{Exemplar-based Agglomerative Clustering}
\description{
Runs exemplar-based agglomerative clustering
}
\usage{
\S4method{aggExCluster}{matrix,missing}(s, x, includeSim=FALSE)
\S4method{aggExCluster}{matrix,ExClust}(s, x, includeSim=FALSE)
\S4method{aggExCluster}{Matrix,missing}(s, x, includeSim=FALSE)
\S4method{aggExCluster}{Matrix,ExClust}(s, x, includeSim=FALSE)
\S4method{aggExCluster}{missing,ExClust}(s, x, includeSim=TRUE)
\S4method{aggExCluster}{function,ANY}(s, x, includeSim=TRUE, ...)
\S4method{aggExCluster}{character,ANY}(s, x, includeSim=TRUE, ...)
}
\arguments{
\item{s}{an \eqn{l\times l}{lxl} similarity matrix or a similarity
function either specified as the name of a package-provided
similarity function as character string or a user provided
function object}
\item{x}{either a prior clustering of class \code{\linkS4class{ExClust}} (or
\code{\linkS4class{APResult}}) or, if called with \code{s} being a
function or function name, input data to be clustered (see
\code{\link{apcluster}} for a detailed specification)}
\item{includeSim}{if \code{TRUE}, the similarity matrix (either computed
internally or passed via the \code{s} argument) is stored to the
slot \code{sim} of the returned
\code{\linkS4class{AggExResult}} object. The default is \code{FALSE}
if \code{aggExCluster} has been called for a similarity matrix,
otherwise the default is \code{TRUE}.}
\item{...}{all other arguments are passed to the selected
similarity function as they are.}
}
\details{\code{aggExCluster} performs agglomerative clustering.
Unlike other methods, e.g., the ones implemented in \code{\link{hclust}},
\code{aggExCluster} is computing exemplars for each cluster and
its merging objective is geared towards the identification of
meaningful exemplars, too.
For each pair of clusters, the merging objective is computed as
follows:
\enumerate{
\item{An intermediate cluster is created as the union
of the two clusters.}
\item{The potential exemplar is selected from the intermediate
cluster as the sample that has the largest average similarity
to all other samples in the intermediate cluster.}
\item{Then the average similarity of the exemplar with all
samples in the first cluster and the average similarity with
all samples in the second cluster is computed. These two values
measure how well the joint exemplar describes the samples in the
two clusters.}
\item{The merging objective is finally computed as the
average of the two measures above. Hence, we can consider the
merging objective as some kind of \dQuote{balanced average
similarity to the joint exemplar}.}
}
In each step, all pairs of clusters are considered and
the pair with the largest merging objective is actually merged.
The joint exemplar is then chosen as the exemplar of the merged
cluster.
\code{aggExCluster} can be used in two ways, either by performing
agglomerative clustering of an entire data set or by performing
agglomerative clustering of data previously clustered by
affinity propagation or another clustering algorithm.
\enumerate{
\item{Agglomerative clustering of an entire data set can be
accomplished either by calling \code{aggExCluster} on a
quadratic similarity matrix without further argument or by
calling \code{aggExCluster} for a function or function name
along with data to be clustered (as argument \code{x}).
A full agglomeration run is performed that starts from \code{l}
clusters (all samples in separate one-element clusters) and ends
with one cluster (all samples in one single cluster).}
\item{Agglomerative clustering starting from a given clustering
result can be accomplished by calling \code{aggExCluster} for
an \code{\linkS4class{APResult}} or \code{\linkS4class{ExClust}}
object passed as parameter \code{x}. The similarity matrix
can either be passed as argument \code{s} or, if missing,
\code{aggExCluster} looks if the similarity matrix is
included in the clustering object \code{x}. A cluster hierarchy
with numbers of clusters ranging from the
number of clusters in \code{x} down to 1 is created.}
}
The result is stored in an \code{\linkS4class{AggExResult}} object.
The slot \code{height} is filled with the merging
objective of each of the \code{maxNoClusters-1} merges. The slot
\code{order} contains a permutation of the samples/clusters for
dendrogram plotting. The algorithm for computing this permutation
is the same as the one used in \code{\link{hclust}}. If \code{aggExCluster}
was called for an entire data set, the slot \code{label}
contains the names of the objects to be clustered (if available,
otherwise the indices are used). If \code{aggExCluster} was called
for a prior clustering, then labels are set to \sQuote{Cluster 1},
\sQuote{Cluster 2}, etc.
}
\note{Similarity matrices can be supplied in dense or sparse
format. Note, however, that sparse matrices are converted to full
dense matrices before clustering which may lead to memory and/or
performance bottlenecks for larger data sets.}
\value{
Upon successful completion, the function returns an
\code{\linkS4class{AggExResult}} object.
}
\author{Ulrich Bodenhofer, Johannes Palme, and Nikola Kostic}
\references{\url{https://github.com/UBod/apcluster}
Bodenhofer, U., Kothmeier, A., and Hochreiter, S. (2011)
APCluster: an R package for affinity propagation clustering.
\emph{Bioinformatics} \bold{27}, 2463-2464.
DOI: \doi{10.1093/bioinformatics/btr406}.
}
\seealso{\code{\linkS4class{AggExResult}}, \code{\link{apcluster-methods}},
\code{\link{plot-methods}}, \code{\link{heatmap-methods}},
\code{\link{cutree-methods}}}
\examples{
## create two Gaussian clouds
cl1 <- cbind(rnorm(50, 0.2, 0.05), rnorm(50, 0.8, 0.06))
cl2 <- cbind(rnorm(50, 0.7, 0.08), rnorm(50, 0.3, 0.05))
x <- rbind(cl1, cl2)
## compute agglomerative clustering from scratch
aggres1 <- aggExCluster(negDistMat(r=2), x)
## show results
show(aggres1)
## plot dendrogram
plot(aggres1)
## plot heatmap along with dendrogram
heatmap(aggres1)
## plot level with two clusters
plot(aggres1, x, k=2)
## run affinity propagation
apres <- apcluster(negDistMat(r=2), x, q=0.7)
## create hierarchy of clusters determined by affinity propagation
aggres2 <- aggExCluster(x=apres)
## show results
show(aggres2)
## plot dendrogram
plot(aggres2)
plot(aggres2, showSamples=TRUE)
## plot heatmap
heatmap(aggres2)
## plot level with two clusters
plot(aggres2, x, k=2)
}
\keyword{cluster}
\keyword{methods}
|
