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
|
\name{mclustICL}
\alias{mclustICL}
\alias{print.mclustICL}
\alias{summary.mclustICL}
\alias{print.summary.mclustICL}
\title{ICL Criterion for Model-Based Clustering}
\description{
ICL (Integrated Complete-data Likelihood) for parameterized Gaussian mixture models fitted by EM algorithm initialized by model-based hierarchical clustering.
}
\usage{
mclustICL(data, G = NULL, modelNames = NULL,
initialization = list(hcPairs = NULL,
subset = NULL,
noise = NULL),
x = NULL, \dots)
\method{summary}{mclustICL}(object, G, modelNames, \dots)
}
\arguments{
\item{data}{
A numeric vector, matrix, or data frame of observations. Categorical
variables are not allowed. If a matrix or data frame, rows
correspond to observations and columns correspond to variables.
}
\item{G}{
An integer vector specifying the numbers of mixture components
(clusters) for which the criteria should be calculated.
The default is \code{G = 1:9}.
}
\item{modelNames}{
A vector of character strings indicating the models to be fitted
in the EM phase of clustering. The help file for
\code{\link{mclustModelNames}} describes the available models.
The default is:
\describe{
\item{\code{c("E", "V")}}{for univariate data}
\item{\code{mclust.options("emModelNames")}}{for multivariate data (n > d)}
\item{\code{c("EII", "VII", "EEI", "EVI", "VEI", "VVI")}}{the spherical and diagonal models for multivariate data (n <= d)}
}
}
\item{initialization}{
A list containing zero or more of the following components:
\describe{
\item{\code{hcPairs}}{
A matrix of merge pairs for hierarchical clustering such as produced
by function \code{hc}. For multivariate data, the default is to compute
a hierarchical clustering tree by applying function \code{hc} with
\code{modelName = "VVV"} to the data or a subset as indicated by the
\code{subset} argument.
The hierarchical clustering results are to start EM.
For univariate data, the default is to use quantiles to start EM.
}
\item{\code{subset}}{
A logical or numeric vector specifying a subset of the data
to be used in the initial hierarchical clustering phase.
}
}
}
\item{x}{
An object of class \code{'mclustICL'}. If supplied, \code{mclustICL}
will use the settings in \code{x} to produce another object of
class \code{'mclustICL'}, but with \code{G} and \code{modelNames}
as specified in the arguments. Models that have already been computed
in \code{x} are not recomputed. All arguments to \code{mclustICL}
except \code{data}, \code{G} and \code{modelName} are
ignored and their values are set as specified in the attributes of
\code{x}.
Defaults for \code{G} and \code{modelNames} are taken from \code{x}.
}
\item{\dots}{
Futher arguments used in the call to \code{\link{Mclust}}.
See also \code{\link{mclustBIC}}.
}
\item{object}{
An integer vector specifying the numbers of mixture components
(clusters) for which the criteria should be calculated.
The default is \code{G = 1:9}.
}
}
\value{
Returns an object of class \code{'mclustICL'} containing the the ICL criterion
for the specified mixture models and numbers of clusters.
The corresponding \code{print} method shows the matrix of values and the top models according to the ICL criterion. The \code{summary} method shows only the top models.
}
\references{
Biernacki, C., Celeux, G., Govaert, G. (2000).
Assessing a mixture model for clustering with the integrated completed likelihood.
\emph{IEEE Trans. Pattern Analysis and Machine Intelligence}, 22 (7), 719-725.
Scrucca L., Fop M., Murphy T. B. and Raftery A. E. (2016) mclust 5: clustering, classification and density estimation using Gaussian finite mixture models, \emph{The R Journal}, 8/1, pp. 289-317.
}
\seealso{
\code{\link{plot.mclustICL}},
\code{\link{Mclust}},
\code{\link{mclustBIC}},
\code{\link{mclustBootstrapLRT}},
\code{\link{bic}},
\code{\link{icl}}
}
\examples{
data(faithful)
faithful.ICL <- mclustICL(faithful)
faithful.ICL
summary(faithful.ICL)
plot(faithful.ICL)
\donttest{
# compare with
faithful.BIC <- mclustBIC(faithful)
faithful.BIC
plot(faithful.BIC)
}
}
\keyword{cluster}
|
