Package 'MixAll'

Description

This package contains methods allowing R users to use the clustering methods of the STK++ library.

Details

As described at the STK++ project's home page, https://www.stkpp.org, STK++ is a versatile, fast, reliable and elegant collection of C++ classes for statistics, clustering, linear algebra, arrays (with an Eigen-like API), regression, dimension reduction, etc. Some functionalities provided by the library are available in the R environment as R functions in MixAll.

The available functionalities are:

1. Clustering (clusterDiagGaussian, clusterCategorical, clusterPoisson, clusterGamma, clusterMixedData)

2. Learning ( (learnDiagGaussian, learnCategorical, learnPoisson, learnGamma, learnMixedData),

3. Prediction (clusterPredict).

Author(s)

Serge Iovleff

Description

Extract parts of a MixAll S4 class

Usage

## S4 method for signature 'ClusterAlgo'
x[i, j, drop]

## S4 replacement method for signature 'ClusterAlgo'
x[i, j] <- value

## S4 method for signature 'ClusterAlgoPredict'
x[i, j, drop]

## S4 replacement method for signature 'ClusterAlgoPredict'
x[i, j] <- value

## S4 method for signature 'ClusterInit'
x[i, j, drop]

## S4 replacement method for signature 'ClusterInit'
x[i, j] <- value

## S4 method for signature 'ClusterStrategy'
x[i, j, drop]

## S4 replacement method for signature 'ClusterStrategy'
x[i, j] <- value

## S4 method for signature 'ClusterCategoricalComponent'
x[i, j, drop]

## S4 method for signature 'ClusterDiagGaussianComponent'
x[i, j, drop]

## S4 method for signature 'ClusterGammaComponent'
x[i, j, drop]

## S4 method for signature 'ClusterPoissonComponent'
x[i, j, drop]

## S4 method for signature 'LearnAlgo'
x[i, j, drop]

## S4 replacement method for signature 'LearnAlgo'
x[i, j] <- value

## S4 method for signature 'KmmComponent'
x[i, j, drop]

Arguments

Description

The data set contains details on the morphology of birds (puffins). Each individual (bird) is described by 6 qualitative variables. One variable for the gender and 5 variables giving a morphological description of the birds. There is 69 puffins divided in 2 sub-classes: lherminieri (34) and subalaris (35).

Format

A data frame with 69 observations on the following 5 variables.

gender

a character vector defining the gender (2 modalities, male or female).

eyebrow

a character vector describing the eyebrow stripe (4 modalities).

collar

a character vector describing the collar (5 modalities).

sub-caudal

a character vector describing the sub-caudal (5 modalities).

border

a character vector describing the border (3 modalities).

Details

This data set is also part of the Rmixmod package.

Source

Bretagnolle, V., 2007. Personal communication, source: Museum.

Examples

data(birds)

Description

Generated data set containing two clusters with untypical ring shapes (circles)

Examples

data(bullsEye)

Description

Generated data set containing two categorical variables for the two clusters with untypical ring shapes (circles)

Examples

data(bullsEye.cat)

Description

Generated data set containing labels for the two clusters with untypical ring shapes (circles)

Examples

data(bullsEye.target)

Description

Car Evaluation Database was derived from a simple hierarchical decision model originally developed for the demonstration of DEX, M. Bohanec, V. Rajkovic: Expert system for decision making.

Format

A data frame with 1728 observations on the following 6 variables.

buying

the buying price (4 modalities: vhigh, high, med, low)

maint

the price of the maintenance (4 modalities: vhigh, high, med, low)

doors

the number of doors (4 modalities: 2, 3, 4, 5more)

persons

the capacity in terms of persons to carry (3 modalities: 2, 4, more)

lug_boot

the size of luggage boot (3 modalities: small, med, big)

safety

the estimated safety of the car (3 modalities: low, med, high)

acceptability

the car acceptability (4 modalities: unacc, acc, good, vgood)

Source

Creator: Marko Bohanec Donors: Marko Bohanec & Blaz Zupan http://archive.ics.uci.edu/ml/datasets/Car+Evaluation

Examples

data(car)

Create an instance of the [ClusterAlgo] class

Description

There is three algorithms and two stopping rules possibles for an algorithm.

• Algorithms:

  • EM: The Expectation Maximisation algorithm

  • CEM: The Classification EM algorithm

  • SEM: The Stochastic EM algorithm

  • SemiSEM: The Semi-Stochastic EM algorithm

• Stopping rules:

  • nbIteration: Set the maximum number of iterations

  • epsilon: Set relative increase of the log-likelihood criterion

• Default values are $200$ nbIteration of EM with an epsilon value of $1.e-8$.

The epsilon value is not used when the algorithm is "SEM" or "SemiSEM".

Usage

clusterAlgo(algo = "EM", nbIteration = 200, epsilon = 1e-07)

Arguments

Value

a [ClusterAlgo] object

Author(s)

Serge Iovleff

Examples

clusterAlgo()
clusterAlgo(algo="SEM", nbIteration=50)
clusterAlgo(algo="CEM", epsilon = 1e-06)

[ClusterAlgo] class for Cluster algorithms.

Description

This class encapsulates the parameters of clustering estimation algorithms methods.

Slots

algo

A character string with the algorithm. Possible values: "SEM", "CEM", "EM", "SemiSEM". Default value: "EM".

nbIteration

Integer defining the maximal number of iterations. Default value: 200.

epsilon

real defining the epsilon value for the algorithm. epsilon is note used if algo is "SEM" or "SemiSEM". Default value: 1e-07.

Examples

getSlots("ClusterAlgo")
new("ClusterAlgo")
new("ClusterAlgo", algo="SEM", nbIteration=1000)

Create an instance of the [ClusterAlgoPredict] class

Description

A prediction algorithm is a two stage algorithm. In the first stage we perform a Monte Carlo algorithm for simulating both missing values and latent class variables. In the second stage, we simulate or impute missing values.

Usage

clusterAlgoPredict(
  algo = "EM",
  nbIterBurn = 50,
  nbIterLong = 100,
  epsilon = 1e-07
)

Arguments

Details

The epsilon value is not used when the algorithm is "SemiSEM".

Value

a [ClusterAlgoPredict] object

Author(s)

Serge Iovleff

Examples

clusterAlgoPredict()
clusterAlgoPredict(algo="SemiSEM", nbIterBurn=0)
clusterAlgoPredict(algo="EM", epsilon = 1e-06)

[ClusterAlgoPredict] class for predict algorithm.

Description

This class encapsulates the parameters of prediction methods.

Slots

algo

A character string with the algorithm. Possible values: "EM", "SemiSEM". Default value: "SemiSEM".

nbIterBurn

Integer defining the number of burning iterations. Default value is 50.

nbIterLong

Integer defining the number of iterations. Default value is 100.

epsilon

real defining the epsilon value for the long algorithm. epsilon is note used if algo is "SemiSEM". Default value: 1e-07.

Examples

getSlots("ClusterAlgoPredict")
new("ClusterAlgoPredict")
new("ClusterAlgoPredict", algo="SemiSEM", nbIterBurn=10)

Create an instance of the [ClusterCategorical] class

Description

This function computes the optimal Categorical mixture model according to the criterion among the list of model given in models and the number of clusters given in nbCluster, using the strategy specified in strategy.

Usage

clusterCategorical(
  data,
  nbCluster = 2,
  models = clusterCategoricalNames(probabilities = "free"),
  strategy = clusterStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [ClusterCategorical] class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the birds data set
data(birds)
## add 10 missing values
x = as.matrix(birds); n <- nrow(x); p <- ncol(x)
indexes <- matrix(c(round(runif(5,1,n)), round(runif(5,1,p))), ncol=2)
x[indexes] <- NA
## estimate model (using fast strategy, results may be misleading)
model <- clusterCategorical( data=x, nbCluster=2:3
                           , models=c( "categorical_pk_pjk", "categorical_p_pjk")
                           , strategy = clusterFastStrategy()
                           )

## use graphics functions

plot(model)


## get summary
summary(model)

## print model (a detailed and very long output)
print(model)

## get estimated missing values
missingValues(model)

Definition of the [ClusterCategorical] class

Description

This class defines a categorical mixture model. It inherits from the [IClusterModel] class. A categorical mixture model is a mixture model of the form

Details

$f({x}|\boldsymbol{\theta}) \\ =\sum_{k=1}^K p_k \prod_{j=1}^d \mathcal{M}(x_j;p_{jk},1) \\ \quad {x} \in \{1,\ldots,L\}^d.$

The probabilities can be assumed equal between all variables in order to reduce the number of parameters.

Slots

component

A [ClusterCategoricalComponent] with the probabilities of the categorical component

Author(s)

Serge Iovleff

Examples

getSlots("ClusterCategorical")
  data(birds)
  new("ClusterCategorical", data=birds)

Definition of the [ClusterCategoricalComponent] class

Description

This class defines a categorical component of a mixture model. It inherits from [IClusterComponent].

Slots

plkj

Array with the probability for the jth variable in the kth cluster to be l.

nbModalities

Integer with the (maximal) number of modalities of the categorical data.

levels

list with the original levels of the variables

Author(s)

Serge Iovleff

See Also

[IClusterComponent] class

Examples

getSlots("ClusterCategoricalComponent")

Description

In a Categorical mixture model, we can build 4 models:

1. The proportions can be equal or free

2. The probabilities can be equal or free for all the variables

Usage

clusterCategoricalNames(prop = "all", probabilities = "all")

clusterValidCategoricalNames(names)

Arguments

Details

The model names are summarized in the following array:

Value

A vector of character with the model names.

Examples

clusterCategoricalNames()
clusterCategoricalNames("all", "equal") # same as c( "categorical_pk_pk", "categorical_p_pk")

Create an instance of the [ClusterDiagGaussian] class

Description

This function computes the optimal diagonal Gaussian mixture model according to the criterion among the list of model given in models and the number of clusters given in nbCluster, using the strategy specified in strategy.

Usage

clusterDiagGaussian(
  data,
  nbCluster = 2,
  models = clusterDiagGaussianNames(),
  strategy = clusterStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [ClusterDiagGaussian] class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the famous geyser data set
data(geyser)
## add 10 missing values as random
x = as.matrix(geyser); n <- nrow(x); p <- ncol(x);
indexes <- matrix(c(round(runif(5,1,n)), round(runif(5,1,p))), ncol=2);
x[indexes] <- NA;
## estimate model (using fast strategy, results may be misleading)
model <- clusterDiagGaussian( data=x, nbCluster=2:3
                            , models=c( "gaussian_pk_sjk")
                            , strategy = clusterFastStrategy()
                            )

## use graphics functions

plot(model)


## get summary
summary(model)

## print model (a detailed and very long output)
print(model)

## get estimated missing values
missingValues(model)

Definition of the [ClusterDiagGaussian] class

Description

This class defines a diagonal Gaussian mixture Model.

Details

This class inherits from the [IClusterModel] class. A diagonal gaussian model is a mixture model of the form:

$f({x}|\boldsymbol{\theta}) =\sum_{k=1}^K p_k \prod_{j=1}^d \phi(x_j;\mu_{jk},\sigma^2_{jk}) \quad x \in {R}^d.$

Some constraints can be added to the variances in order to reduce the number of parameters.

Slots

component

A [ClusterDiagGaussianComponent] with the mean and standard deviation of the diagonal mixture model.

Author(s)

Serge Iovleff

See Also

[IClusterModel] class

Examples

getSlots("ClusterDiagGaussian")
data(geyser)
new("ClusterDiagGaussian", data=geyser)

Definition of the [ClusterDiagGaussianComponent] class

Description

This class defines a diagonal Gaussian component of a mixture model. It inherits from [IClusterComponent].

Slots

mean

Matrix with the mean of the jth variable in the kth cluster.

sigma

Matrix with the standard deviation of the jth variable in the kth cluster.

Author(s)

Serge Iovleff

See Also

[IClusterComponent] class

Examples

getSlots("ClusterDiagGaussianComponent")

Description

In a diagonal Gaussian mixture model, we assume that the variance matrices are diagonal in each cluster. Assumptions on the proportions and standard deviations give rise to 8 models:

1. The proportions can be equal or free

2. The standard deviations can be equal or free for all the variables

3. The standard deviations can be equal or free for all the clusters

Usage

clusterDiagGaussianNames(
  prop = "all",
  sdInCluster = "all",
  sdBetweenCluster = "all"
)

clusterValidDiagGaussianNames(names)

Arguments

Details

The model names are summarized in the following array:

Value

A vector of character with the model names.

Examples

clusterDiagGaussianNames()
## same as c("gaussian_p_sk", "gaussian_pk_sk")
clusterDiagGaussianNames(prop="all", sdInCluster="equal", sdBetweenCluster= "free")

Create an instance of the [ClusterGamma] class

Description

This function computes the optimal gamma mixture model according to the criterion among the list of model given in models and the number of clusters given in nbCluster, using the strategy specified in strategy.

Usage

clusterGamma(
  data,
  nbCluster = 2,
  models = "gamma_pk_ajk_bjk",
  strategy = clusterStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [ClusterGamma] class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the famous geyser data set
data(geyser)
## add 10 missing values
x = geyser;
x[round(runif(5,1,nrow(geyser))), 1] <- NA
x[round(runif(5,1,nrow(geyser))), 2] <- NA

## use graphics functions
set.seed(2)
model <- clusterGamma( data=x, nbCluster=2:3
                     , models="gamma_pk_ajk_bjk"
                     , strategy = clusterFastStrategy())

## use plot

plot(model)


## get summary
summary(model)

## print model (a detailed and very long output)
print(model)

## get estimated missing values
missingValues(model)

Definition of the [ClusterGamma] class

Description

This class inherits from the [IClusterModel] class. A gamma mixture model is a mixture model of the form:

$f({x}|\boldsymbol{\theta}) \\ =\sum_{k=1}^K p_k \prod_{j=1}^d \gamma(x_j;a_{jk},b_{jk}) \\ \quad {x} \in {R}^d.$

Constraints can be added to the shapes and/or scales in order to reduce the number of parameters.

Slots

component

A [ClusterGammaComponent] with the shapes and the scales of the component mixture model.

Author(s)

Serge Iovleff

See Also

[IClusterModel] class

Examples

getSlots("ClusterGamma")
  data(geyser)
  new("ClusterGamma", data=geyser)

Definition of the [ClusterGammaComponent] class

Description

This class defines a gamma component of a mixture Model. It inherits from [IClusterComponent].

Slots

shape

Matrix with the shapes of the jth variable in the kth cluster.

scale

Matrix with the scales of the jth variable in the kth cluster.

Author(s)

Serge Iovleff

See Also

[IClusterComponent] class

Examples

getSlots("ClusterGammaComponent")

Description

In a gamma mixture model, we can assume that the shapes are equal in each/all cluster(s) or not. We can also assume that the scales are equal in each/all cluster(s) or not.

Usage

clusterGammaNames(
  prop = "all",
  shapeInCluster = "all",
  shapeBetweenCluster = "all",
  scaleInCluster = "all",
  scaleBetweenCluster = "all"
)

clusterValidGammaNames(names)

Arguments

Details

Some configuration are impossibles. If the shapes are equal between all the clusters, then the scales cannot be equal between all the clusters. Conversely if the scales are equal between all the cluster, then the shapes cannot be equal between all the clusters.

This gives rise to 24 models:

1. The proportions can be equal or free

2. The shapes can be equal or free in each clusters

3. The shapes can be equal or free between all clusters

4. The scales can be equal or free for each clusters

5. The scales can be equal or free between all clusters

The model names are summarized in the following array:

Value

A vector of character with the model names.

Examples

clusterGammaNames()
## same as c("gamma_p_ak_bj", "gamma_pk_ak_bj")
clusterGammaNames("all", "equal", "free", "free", "equal")

Create an instance of [ClusterInit] class

Description

The initialization step is a two stages process: the proper initialization step and some (optionnals) iterations of an algorithm [clusterAlgo].

Usage

clusterInit(
  method = "class",
  nbInit = 5,
  algo = "EM",
  nbIteration = 20,
  epsilon = 0.01
)

Arguments

Details

There is three ways to initialize the parameters:

• random: The initial parameters of the mixture are chosen randomly

• class: The initial membership of individuals are sampled randomly

• fuzzy: The initial probabilities of membership of individuals are sampled randomly

A few iterations of an algorithm [clusterAlgo] are then performed. It is strongly recommended to use a few number of iterations of the EM or SEM algorithms after initialization. This allows to detect "bad" initialization starting point.

These two stages are repeated until nbInit is reached. The initial point with the best log-likelihood is conserved as the initial starting point.

Value

a [ClusterInit] object

Author(s)

Serge Iovleff

Examples

clusterInit(method = "class", nbInit=1, algo="CEM",nbIteration=50, epsilon=0.00001)
 clusterInit(nbIteration=0) # no algorithm

Constructor of the [ClusterInit] class

Description

This class encapsulates the parameters of clustering initialization methods.

Slots

method

Character string with the initialization method to use. Default value: "class"

nbInit

Integer defining the number of initialization to perform. Default value: 5.

algo

An instance of ClusterAlgo class. Default value: clusterAlgo("EM", 20, 0.01).

Author(s)

Serge Iovleff

Examples

getSlots("ClusterInit")
  new("ClusterInit")
  new("ClusterInit", nbInit=1)

Create an instance of the [ClusterMixedDataModel] class

Description

This function computes the optimal mixture model for mixed data according to the criterion among the number of clusters given in nbCluster using the strategy specified in [strategy].

Usage

clusterMixedData(
  data,
  models,
  nbCluster = 2,
  strategy = clusterStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [ClusterMixedDataModel] class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the heart disease data set
data(HeartDisease.cat)
data(HeartDisease.cont)
## with default values
ldata = list(HeartDisease.cat, HeartDisease.cont);
models = c("categorical_pk_pjk","gaussian_pk_sjk")
model <- clusterMixedData(ldata, models, nbCluster=2:5, strategy = clusterFastStrategy())

## get summary
summary(model)

## get estimated missing values
missingValues(model)


## print model (a very detailed output)
print(model)
## use graphics functions
plot(model)

Definition of the [ClusterMixedDataModel] class

Description

This class defines a mixed data mixture Model.

Details

This class inherits from the [IClusterModel] class. A model for mixed data is a mixture model of the form:

$f({{x}}_i=({{x}}_{1i}, {{x}}_{2i},\ldots {{x}}_{Li})|\theta) = \sum_{k=1}^K p_k \prod_{l=1}^L h({{x}}_{li}| \lambda_{lk},\alpha_l).$

The density functions (or probability distribution functions)

$h(.|\lambda_{lk},\alpha_l)$

can be any implemented model (Gaussian, Poisson,...).

Slots

lcomponent

a list of [IClusterComponent]

Author(s)

Serge Iovleff

See Also

[IClusterModel] class

Examples

getSlots("ClusterMixedDataModel")

Create an instance of the [ClusterPoisson] class

Description

This function computes the optimal poisson mixture model according to the [criterion] among the list of model given in [models] and the number of clusters given in [nbCluster], using the strategy specified in [strategy].

Usage

clusterPoisson(
  data,
  nbCluster = 2,
  models = clusterPoissonNames(),
  strategy = clusterStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [ClusterPoisson] class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the DebTrivedi data set.
data(DebTrivedi)
dt <- DebTrivedi[1:500, c(1, 6,8, 15)]

model <- clusterPoisson( data=dt, nbCluster=2
                       , models=clusterPoissonNames(prop = "equal")
                       , strategy = clusterFastStrategy())

## use graphics functions

plot(model)


## get summary
summary(model)

## print model (a very detailed output)
print(model)

## get estimated missing values
missingValues(model)

Definition of the [ClusterPoisson] class

Description

This class inherits from the [IClusterModel] class. A poisson mixture model is a mixture model of the form:

$f({x}|\boldsymbol{\theta}) \\ =\sum_{k=1}^K p_k \prod_{j=1}^d Pois(x_j;\lambda_{jk}) \\ \quad {x} \in {N}^d.$

Slots

component

A [ClusterPoissonComponent] with the lambda of the component mixture model.

Author(s)

Serge Iovleff

See Also

[IClusterModel] class

Examples

getSlots("ClusterPoisson")
  data(DebTrivedi)
  dt <- DebTrivedi[, c(1, 6,8, 15)]
  new("ClusterPoisson", data=dt)

Description

In a Poisson mixture model, we can build 4 models:

1. The proportions can be equal or free

2. The means can be equal, free or proportional for all the variables

Usage

clusterPoissonNames(prop = "all", mean = "all")

clusterValidPoissonNames(names)

Arguments

Details

The model names are summarized in the following array:

Value

A vector of character with the model names.

Examples

clusterPoissonNames()
clusterPoissonNames("all", "proportional") # same as c( "poisson_pk_ljlk", "poisson_p_ljlk")

Create an instance of [ClusterPredict] class

Description

This function predicts the best cluster each sample in data belongs to.

Usage

clusterPredict(data, model, algo = clusterAlgoPredict(), nbCore = 1)

Arguments

Value

An instance of [ClusterPredict] with predicted values

Author(s)

Serge Iovleff

Examples

## A quantitative example with the famous iris data set
data(iris)
## get quantitatives 
x = as.matrix(iris[1:4])
## sample train and test data sets
indexes <- sample(1:nrow(x), nrow(x)/2)
train <- x[ indexes,]
test  <- x[-indexes,]
## estimate model (using fast strategy, results may be misleading)
model1 <- clusterDiagGaussian( data =train, nbCluster=2:3
                             , models=c( "gaussian_p_sjk")
                             )
## get summary
summary(model1)
## compute prediction and compare
model2 <- clusterPredict(test, model1)
show(model2)
as.integer(iris$Species[-indexes])

Class [ClusterPredict] for predicting

Description

This class encapsulate the parameters for predicted data.

Slots

data

Matrix with the data set

missing

Matrix with the indexes of the missing values

Author(s)

Serge Iovleff

See Also

[IClusterPredict] class

Examples

getSlots("ClusterPredict")

Class [ClusterPredictMixedData] for predicting

Description

This class encapsulate the parameters for predicted data.

Slots

ldata

list of matrix with the data sets

lmissing

list of matrix with the indexes of the missing values

Author(s)

Serge Iovleff

See Also

[IClusterPredict] class

Examples

getSlots("ClusterPredictMixedData")

Description

A strategy is a way to find a good estimate of the parameters of a mixture model when using an EM algorithm or its variants. A “try” is composed of three stages

• nbShortRun short iterations of the initialization step and of the EM, CEM, SEM or SemiSEM algorithm.

• nbInit initializations using the [clusterInit] method.

• A long run of the EM, CEM, SEM or SemiSEM algorithm.

For example if nbInit is 5 and nbShortRun is also 5, there will be 5 packets of 5 models initialized. In each packet, the best model will be ameliorated using a short run. Among the 5 models ameliorated the best one will be estimated until convergence using a long run. In total there will be 25 initializations, 5 short runs and one long-run.

clusterSemiSEMStrategy() create an instance of [ClusterStrategy] for users with many missing values uning a semiSem algorithm.

clusterSEMStrategy() create an instance of [ClusterStrategy] for users with many missing values using a SEM algorithm.

clusterFastStrategy() create an instance of [ClusterStrategy] for impatient user.

Usage

clusterStrategy(
  nbTry = 1,
  nbInit = 5,
  initMethod = "class",
  initAlgo = "EM",
  nbInitIteration = 20,
  initEpsilon = 0.01,
  nbShortRun = 5,
  shortRunAlgo = "EM",
  nbShortIteration = 100,
  shortEpsilon = 1e-04,
  longRunAlgo = "EM",
  nbLongIteration = 1000,
  longEpsilon = 1e-07
)

clusterSemiSEMStrategy()

clusterSEMStrategy()

clusterFastStrategy()

Arguments

Details

The whole process can be repeated at least nbTry times. If a try success, the estimated model is returned, otherwise an empty model is returned (with an error message).

Value

a [ClusterStrategy] object

Author(s)

Serge Iovleff

Examples

clusterStrategy()
   clusterStrategy(longRunAlgo= "CEM", nbLongIteration=100)
   clusterStrategy(nbTry = 1, nbInit= 1, shortRunAlgo= "SEM", nbShortIteration=100)

   clusterSemiSEMStrategy()

   clusterSEMStrategy()

   clusterFastStrategy()

Constructor of [ClusterStrategy] class

Description

This class encapsulate the parameters of the clustering estimation strategies.

Details

@slot nbTry Integer defining the number of tries. Default value: 1. @slot nbShortRun Integer defining the number of short run. Recall that the strategy launch an initialization before each short run. Default value is 5. @slot initMethod A [ClusterInit] object defining the way to initialize the estimation method. Default value is [ClusterInit]. @slot shortAlgo A [ClusterAlgo] object defining the algorithm to use during the short runs of the estimation method. Default value is clusterAlgo("EM",100,1e-04). @slot longAlgo A [ClusterAlgo] object defining the algorithm to use during the long run of the estimation method. Default value is clusterAlgo("EM",1000,1e-07).

Author(s)

Serge Iovleff

Examples

new("ClusterStrategy")
  shortAlgo=clusterAlgo("SEM",1000)
  longAlgo =clusterAlgo("SemiSEM",200,1e-07)
  new("ClusterStrategy", shortAlgo=shortAlgo, longAlgo=longAlgo)
  getSlots("ClusterStrategy")

Description

Deb and Trivedi (1997) analyze data on 4406 individuals, aged 66 and over, who are covered by Medicare, a public insurance program. Originally obtained from the US National Medical Expenditure Survey (NMES) for 1987/88, the data are available from the data archive of the Journal of Applied Econometrics. It was prepared for an R package accompanying Kleiber and Zeileis (2008) and is also available asDebTrivedi.rda in the Journal of Statistical Software together with Zeileis (2006). The objective is to model the demand for medical care -as captured by the number of physician/non-physician office and hospital outpatient visits- by the covariates available for the patients.

Source

https://www.jstatsoft.org/htaccess.php?volume=27&type=i&issue=08&filename=paper

References

Zeileis, A. and Kleiber, C. and Jackma, S. (2008). "Regression Models for Count Data in R". JSS 27, 8, 1–25.

Examples

data(DebTrivedi)

Description

The file geyser.rda contains 272 observations from the Old Faithful Geyser in the Yellowstone National Park. Each observation consists of two measurements: the duration (in minutes) of the eruption and the waiting time (in minutes) to the next eruption.

Format

A data frame with 272 observations on the following 2 variables.

Duration

a numeric vector containing the duration (in minutes) of the eruption

Waiting.Time

a numeric vector containing the waiting time (in minutes) to the next eruption

Details

Old Faithful erupts more frequently than any other big geyser, although it is not the largest nor the most regular geyser in the park. Its average interval between two eruptions is about 76 minutes, varying from 45 - 110 minutes. An eruption lasts from 1.1/2 to 5 minutes, expels 3,700 - 8,400 gallons (14,000 - 32,000 liters) of boiling water, and reaches heights of 106 - 184 feet (30 - 55m). It was named for its consistent performance by members of the Washburn Expedition in 1870. Old Faithful is still as spectacular and predictable as it was a century ago.

Source

https://web.archive.org/web/20191110083004/http://www.geyserstudy.org/geyser.aspx?pGeyserNo=OLDFAITHFUL

References

Hardle, W. (1991). "Smoothing Techniques with Implementation in S". Springer-Verlag, New York. Azzalini, A. and Bowman, A. W. (1990). "A look at some data on the Old Faithful geyser". Applied Statistics 39, 357-365.

Examples

data(geyser)

Description

The Cleveland Heart Disease Data found in the UCI machine learning repository consists of 14 variables measured on 303 individuals who have heart disease. The individuals had been grouped into five levels of heart disease. The information about the disease status is in the HeartDisease.target data set.

Format

Three data frames with 303 observations on the following 14 variables.

age

age in years

sex

sex (1 = male; 0 = female)

cp

chest pain type. 1: typical angina, 2: atypical angina, 3: non-anginal pain, 4: asymptomatic

trestbps

resting blood pressure (in mm Hg on admission to the hospital)

chol

serum cholestoral in mg/dl

fbs

(fasting blood sugar > 120 mg/dl) (1 = true; 0 = false)

restecg

resting electrocardiographic results. 0: normal, 1: having ST-T wave abnormality (T wave inversions and/or ST, elevation or depression of > 0.05 mV) 2: showing probable or definite left ventricular hypertrophy by Estes\' criteria

thalach

maximum heart rate achieved

exang

exercise induced angina (1 = yes; 0 = no)

oldpeak

ST depression induced by exercise relative to rest

slope

the slope of the peak exercise ST segment 1: upsloping, 2: flat, 3: downsloping

ca

number of major vessels (0-3) colored by flourosopy (4 missing values)

thal

3 = normal; 6 = fixed defect; 7 = reversable defect (2 missing values)

num

diagnosis of heart disease (angiographic disease status). 0: < 50 1: > 50 (in any major vessel: attributes 59 through 68 are vessels)

Details

The variables consist of five continuous and eight discrete attributes, the former in the HeartDisease.cont data set and the later in the HeartDisease.cat data set. Three of the discrete attributes have two levels, three have three levels and two have four levels. There are six missing values in the data set.

Source

Author: David W. Aha (aha 'AT' ics.uci.edu) (714) 856-8779

Donors: The data was collected from the Cleveland Clinic Foundation (cleveland.data)

https://archive.ics.uci.edu/ml/datasets/Heart+Disease

Detrano, R., Janosi, A., Steinbrunn, W., Pfisterer, M., Schmid, J., Sandhu, S., Guppy, K., Lee, S., & Froelicher, V. (1989). International application of a new probability algorithm for the diagnosis of coronary artery disease. American Journal of Cardiology, 64,304–310.

David W. Aha & Dennis Kibler. "Instance-based prediction of heart-disease presence with the Cleveland database."

Gennari, J.H., Langley, P, & Fisher, D. (1989). Models of incremental concept formation. Artificial Intelligence, 40, 11–61.

Examples

summary(data(HeartDisease.cat))
summary(data(HeartDisease.cont))
summary(data(HeartDisease.target))

Definition of the [IClusterComponent] class

Description

Interface base class defining a component of a mixture Model

This class defines a poisson component of a mixture Model. It inherits from [IClusterComponent].

Slots

data

Matrix with the data set

missing

Matrix with the indexes of the missing values

modelName

model name associated with the data set

lambda

Matrix with the mean of the jth variable in the kth cluster.

Author(s)

Serge Iovleff

See Also

[IClusterComponent] class

Examples

getSlots("IClusterComponent")

getSlots("ClusterPoissonComponent")

Interface base Class [IClusterModel] for Cluster models.

Description

This class encapsulate the common parameters of all the Cluster models.

Details

A Cluster model is a model of the form

$f({x}|\boldsymbol{\theta}) \sum_{k=1}^K p_k h({x};\boldsymbol{\lambda}_k,\boldsymbol{\alpha}) \quad {x} \in J.$

where h can be either a pdf, a discrete probability, (homogeneous case) or a product of arbitrary pdf and discrete probabilities (mixed data case).

Slots

nbSample

Integer with the number of samples of the model.

nbCluster

Integer with the number of cluster of the model.

pk

Vector of size K with the proportions of each mixture.

tik

Matrix of size $n \times K$ with the posterior probability of the ith individual to belong to kth cluster.

lnFi

Vector of size n with the log-likelihood of the ith individuals.

zi

Vector of integer of size n with the attributed class label of the individuals.

ziFit

Vector of integer of size n with the fitted class label of the individuals (only used in supervised learning).

lnLikelihood

Real given the ln-liklihood of the Cluster model.

criterion

Real given the value of the AIC, BIC, ICL or ML criterion.

criterionName

string with the name of the criterion. Possible values are "BIC", "AIC", "ICL" or "ML". Default is "ICL".

nbFreeParameter

Integer given the number of free parameters of the model.

strategy

the instance of the [ClusterStrategy] used in the estimation process of the mixture. Default is clusterStrategy().

Author(s)

Serge Iovleff

Examples

getSlots("IClusterModel")

Interface class [IClusterPredict] for predicting

Description

Interface base class for predicting clusters

Slots

nbSample

Integer with the number of samples

nbCluster

Integer with the number of cluster

pk

Vector of size K with the proportions of each mixture.

tik

Matrix of size $n \times K$ with the posterior probability of the ith individual to belong to kth cluster.

lnFi

Vector of size n with the log-likelihood of the ith individuals.

zi

Vector of integer of size n with the attributed class label of the individuals

algo

an instance of [ClusterAlgoPredict]

model

an instance of a (derived) [IClusterModel]

Author(s)

Serge Iovleff

Examples

getSlots("IClusterPredict")

Create an instance of the [KmmModel] class

Description

This function computes the optimal kernel mixture model (KMM) according to the [criterion] among the number of clusters given in [nbCluster], using the strategy specified in [strategy].

Usage

kmm(
  data,
  nbCluster = 2,
  dim = 10,
  models = "kmm_pk_s",
  kernelName = "Gaussian",
  kernelParameters = c(1),
  kernelComputation = TRUE,
  strategy = kmmStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [KmmModel] class.

Note

in KmmModel instance returned, the gram matrix is computed if and only if kernelComputation is TRUE.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the famous bulls eye model
data(bullsEye)
## estimate model
model <- kmm( data=bullsEye, nbCluster=2:3, models= "kmm_pk_s")


## get summary
summary(model)
## use graphics functions

plot(model)

Definition of the [KmmComponent] class

Description

This class defines a kernel component of a mixture Model. It inherits from [IClusterComponent].

Slots

dim

Vector with the dimension of the kth cluster

sigma2

Vector with the standard deviation in the kth cluster.

gram

Matrix storing the gram matrix if its computation is needed

kernelName

string with the name of the kernel to use. Possible values: "Gaussian", "polynomial", "Laplace", "linear","rationalQuadratic", "Hamming". Default is "Gaussian".

kernelParameters

vector with the parameters of the kernel.

kernelComputation

boolean value set as TRUE if Gram matrix is to be computed FALSE othewise. Default is TRUE.

Author(s)

Serge Iovleff

See Also

[IClusterComponent] class

Examples

getSlots("KmmComponent")

Create an instance of the [KmmMixedDataModel] class

Description

This function computes the optimal mixture model for mixed data using kernel mixture models according to the criterion among the number of clusters given in nbCluster using the strategy specified in [strategy].

Usage

kmmMixedData(
  ldata,
  lmodels,
  nbCluster = 2,
  strategy = clusterStrategy(),
  criterion = "ICL",
  nbCore = 1
)

Arguments

Details

For each data set in data, we need to specify a list of parameters

Value

An instance of the [KmmMixedDataModel] class.

Author(s)

Serge Iovleff

Examples

## An example with the bullsEye data set
data(bullsEye)
data(bullsEye.cat)
## with default values
ldata     <- list(bullsEye, bullsEye.cat)
modelcont <- list(modelName="kmm_pk_s", dim = 10, kernelName="Gaussian")
modelcat  <- list(modelName="kmm_pk_s", dim = 20, kernelName="Hamming", kernelParameters = c(0.6))
lmodels   <- list( modelcont, modelcat)

model <- kmmMixedData(ldata, lmodels, nbCluster=2:5, strategy = clusterFastStrategy())

## get summary
summary(model)



## use graphics functions
plot(model)

Definition of the [KmmMixedDataModel] class

Description

This class defines a mixed data kernel mixture Model (KMM).

Details

This class inherits from the [IClusterModel] class. A model for mixed data is a mixture model of the form:

$f({{x}}_i=({{x}}_{1i}, {{x}}_{2i},\ldots {{x}}_{Li})|\theta) = \sum_{k=1}^K p_k \prod_{l=1}^L h({{x}}_{li}).$

The density functions (or probability distribution functions)

$h(.)$

can be any implemented kmm model on a RKHS space.

Slots

lcomponent

a list of [KmmComponent]

Author(s)

Serge Iovleff

See Also

[IClusterModel] class

Examples

getSlots("KmmMixedDataModel")

Definition of the [KmmModel] class

Description

This class defines a Kernel mixture Model (KMM).

Details

This class inherits from the [IClusterModel] virtual class. A KMM is a mixture model of the form:

$f({x}|\boldsymbol{\theta}) =\sum_{k=1}^K p_k \prod_{j=1}^d \phi(x_j;\sigma^2_{k}) \quad x \in {R}^d.$

Some constraints can be added to the variances in order to reduce the number of parameters.

Slots

component

A [KmmComponent] with the dimension and standard deviation of the kernel mixture model.

Author(s)

Serge Iovleff

See Also

[IClusterModel] class

Examples

getSlots("KmmModel")
data(bullsEye)
new("KmmModel", data=bullsEye)

Description

In a Kernel mixture model, sssumptions on the proportions and standard deviations give rise to 4 models:

1. Proportions can be equal or free.

2. Standard deviations are equal or free for all clusters.

Usage

kmmNames(prop = "all", sdBetweenCluster = "all")

kmmValidModelNames(names)

kmmValidKernelNames(names)

Arguments

Details

The model names are summarized in the following array:

Value

A vector of character with the model names.

TRUE if the names in the vector names are valid, FALSE otherwise.

Examples

kmmNames()
## same as c("kmm_p_sk")
kmmNames( prop = "equal", sdBetweenCluster= "free")

Create an instance of [ClusterStrategy] class

Description

A strategy is a multistage empirical process for finding a good estimate in the clustering estimation process.

Usage

kmmStrategy(
  nbTry = 1,
  nbInit = 5,
  initMethod = "class",
  initAlgo = "EM",
  nbInitIteration = 20,
  initEpsilon = 0.01,
  nbShortRun = 5,
  shortRunAlgo = "EM",
  nbShortIteration = 100,
  shortEpsilon = 1e-04,
  longRunAlgo = "EM",
  nbLongIteration = 1000,
  longEpsilon = 1e-07
)

Arguments

Details

A strategy is a way to find a good estimate of the parameters of a kernel mixture model when using an EM algorithm or its variants. A “try” of kmmStrategy is composed of three stages

• nbShortRun short iterations of the initialization step and of the EM, CEM or SEM algorithm.

• nbInit initializations using the [clusterInit] method.

• A long run of the EM, CEM or SEM algorithm.

For example if nbInit is 5 and nbShortRun is also 5, there will be 5 times 5 models initialized. Five time, the best model (in the likelihood sense) will be ameliorated using a short run. Among the 5 models ameliorated one will be estimated until convergence using a long run. In total there is 25 initializations.

The whole process can be repeated at least nbTry times. If a try success, the estimated model is returned, otherwise an empty model is returned.

Value

a [ClusterStrategy] object

Author(s)

Serge Iovleff

Examples

kmmStrategy()
   kmmStrategy(longRunAlgo= "CEM", nbLongIteration=100)
   kmmStrategy(nbTry = 1, nbInit= 1, shortRunAlgo= "EM", nbShortIteration=100)

Create an instance of the [LearnAlgo] class

Description

There is two algorithms and two stopping rules possibles for a learning algorithm.

• Algorithms:

  • Impute: Impute the missing values during the iterations

  • Simul: Simulate the missing values during the iterations

• Stopping rules:

  • nbIteration: Set the maximum number of iterations

  • epsilon: Set relative increase of the log-likelihood criterion

• Default values are $200$ nbIteration of Simul.

The epsilon value is not used when the algorithm is "Simul". It is worth noting that if there is no missing values, the method should be "Impute" and nbIteration should be set to 1!

Usage

learnAlgo(algo = "Simul", nbIteration = 200, epsilon = 1e-07)

Arguments

Value

a [LearnAlgo] object

Author(s)

Serge Iovleff

Examples

learnAlgo()
learnAlgo(algo="simul", nbIteration=50)
learnAlgo(algo="impute", epsilon = 1e-06)

[LearnAlgo] class for Cluster algorithms.

Description

This class encapsulates the parameters of clustering estimation algorithms methods.

Slots

algo

A character string with the algorithm. Possible values: "Simul", "Impute. Default value: "Simul".

nbIteration

Integer defining the maximal number of iterations. Default value: 200.

epsilon

real defining the epsilon value for the algorithm. epsilon is note used if algo is "Simul". Default value: 1e-07.

Examples

getSlots("LearnAlgo")
new("LearnAlgo")
new("LearnAlgo", algo="Impute", nbIteration=100)

Description

This function learn the optimal mixture model when the class labels are known according to the criterion among the list of model given in models.

Usage

learnDiagGaussian(
  data,
  labels,
  prop = NULL,
  models = clusterDiagGaussianNames(prop = "equal"),
  algo = "simul",
  nbIter = 100,
  epsilon = 1e-08,
  criterion = "ICL",
  nbCore = 1
)

learnPoisson(
  data,
  labels,
  prop = NULL,
  models = clusterPoissonNames(prop = "equal"),
  algo = "simul",
  nbIter = 100,
  epsilon = 1e-08,
  criterion = "ICL",
  nbCore = 1
)

learnGamma(
  data,
  labels,
  prop = NULL,
  models = clusterGammaNames(prop = "equal"),
  algo = "simul",
  nbIter = 100,
  epsilon = 1e-08,
  criterion = "ICL",
  nbCore = 1
)

learnCategorical(
  data,
  labels,
  prop = NULL,
  models = clusterCategoricalNames(prop = "equal"),
  algo = "simul",
  nbIter = 100,
  epsilon = 1e-08,
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of a learned mixture model class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the famous iris data set
data(iris)

## get data and target
x <- as.matrix(iris[,1:4]);
z <- as.vector(iris[,5]);
n <- nrow(x); p <- ncol(x);

## add missing values at random
indexes <- matrix(c(round(runif(5,1,n)), round(runif(5,1,p))), ncol=2);
x[indexes] <- NA;

## learn model
model <- learnDiagGaussian( data=x, labels= z, prop = c(1/3,1/3,1/3)
                          , models = clusterDiagGaussianNames(prop = "equal")
                          )

## get summary
summary(model)

## use graphics functions

plot(model)



## print model (a detailed and very long output)
print(model)


## get estimated missing values
missingValues(model)

Description

This function learn the optimal mixture model when the class labels are known according to the criterion among the list of model given in models.

Usage

learnMixedData(
  data,
  models,
  labels,
  prop = NULL,
  algo = "impute",
  nbIter = 100,
  epsilon = 1e-08,
  criterion = "ICL",
  nbCore = 1
)

Arguments

Value

An instance of the [ClusterMixedDataModel] class.

Author(s)

Serge Iovleff

Examples

## A quantitative example with the heart disease data set
data(HeartDisease.cat)
data(HeartDisease.cont)
## with default values
ldata = list(HeartDisease.cat, HeartDisease.cont);
models = c("categorical_pk_pjk","gaussian_pk_sjk")
model <- clusterMixedData(ldata, models, nbCluster=2:5, strategy = clusterFastStrategy())

## get summary
summary(model)

## get estimated missing values
missingValues(model)


## print model (a detailed and very long output)
print(model)
## use graphics functions
plot(model)

Description

The missing methods allow the user to get the imputed mssing values from a mixture model.

Usage

missingValues(x)

## S4 method for signature 'ClusterMixedDataModel'
missingValues(x)

## S4 method for signature 'ClusterDiagGaussianComponent'
missingValues(x)

## S4 method for signature 'ClusterDiagGaussian'
missingValues(x)

## S4 method for signature 'ClusterGammaComponent'
missingValues(x)

## S4 method for signature 'ClusterGamma'
missingValues(x)

## S4 method for signature 'ClusterCategoricalComponent'
missingValues(x)

## S4 method for signature 'ClusterCategorical'
missingValues(x)

## S4 method for signature 'ClusterPoissonComponent'
missingValues(x)

## S4 method for signature 'ClusterPoisson'
missingValues(x)

## S4 method for signature 'ClusterPredict'
missingValues(x)

## S4 method for signature 'ClusterPredictMixedData'
missingValues(x)

## S4 method for signature 'KmmComponent'
missingValues(x)

## S4 method for signature 'KmmModel'
missingValues(x)

Arguments

Value

A matrix with three columns (row index, column index, value)

Examples

## add 10 missing values as random
data(geyser)
x = as.matrix(geyser); n <- nrow(x); p <- ncol(x);
indexes <- matrix(c(round(runif(5,1,n)), round(runif(5,1,p))), ncol=2);
x[indexes] <- NA;
## estimate model (using fast strategy, results may be misleading)
model <- clusterDiagGaussian(data=x, nbCluster=2:3, strategy = clusterFastStrategy())
missingValues(model)

Plotting of a class [ClusterCategorical]

Description

Plotting data from a [ClusterCategorical] object using the estimated parameters and partition.

Usage

## S4 method for signature 'ClusterCategorical'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## the car data set (verify car data is in your environment)
  data(car)
  model <- clusterCategorical(car, 3, strategy = clusterFastStrategy())
  plot(model)

Plotting of a class [ClusterDiagGaussian]

Description

Plotting data from a [ClusterDiagGaussian] object using the estimated parameters and partition.

Usage

## S4 method for signature 'ClusterDiagGaussian'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## the famous iris data set
  data(iris)
  model <- clusterDiagGaussian(iris[1:4], 3, strategy = clusterFastStrategy())
  plot(model)
  plot(model, c(1,3))
  plot(model, c("Sepal.Length","Sepal.Width"))

Plotting of a class [ClusterGamma]

Description

Plotting data from a [ClusterGamma] object using the estimated parameters and partition.

Usage

## S4 method for signature 'ClusterGamma'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## Example with quantitative vairables
  data(iris)
  model <- clusterGamma( data=iris[1:4], nbCluster=3
                       , models=clusterGammaNames(prop = "equal")
                       , strategy = clusterFastStrategy())
  plot(model)
  plot(model, c(1,3))
  plot(model, c("Sepal.Length","Sepal.Width"))

Plotting of a class [ClusterMixedDataModel]

Description

Plotting data from a [ClusterMixedDataModel] object using the estimated parameters and partition.

Usage

## S4 method for signature 'ClusterMixedDataModel'
plot(x, y, ...)

Arguments

See Also

plot

Plotting of a class [ClusterPoisson]

Description

Plotting data from a [ClusterPoisson] object using the estimated parameters and partition.

Usage

## S4 method for signature 'ClusterPoisson'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## Example with counting data
  data(DebTrivedi)
  dt <- DebTrivedi[, c(1, 6,8, 15)]
  model <- clusterPoisson(iris[1:4], 3, strategy = clusterFastStrategy())
  plot(model)
  plot(model, c(1,2))

Plotting of a class [KmmComponent]

Description

Plotting data from a [KmmComponent] object using the estimated partition.

Usage

## S4 method for signature 'KmmComponent'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## the bull eyes data set
  data(bullsEye)
  model <- kmm( bullsEye, 2, models= "kmm_pk_s")
  plot(model)

Plotting of a class [KmmMixedDataModel]

Description

Plotting data from a [KmmMixedDataModel] object using the estimated parameters and partition.

Usage

## S4 method for signature 'KmmMixedDataModel'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## The bullsEye data set
data(bullsEye)
data(bullsEye.cat)
## with default values
ldata  = list(bullsEye, bullsEye.cat)
modelcont <- list(modelName="kmm_pk_s", dim = 10, kernelName="Gaussian")
modelcat  <- list(modelName="kmm_pk_s", dim = 20, kernelName="Hamming", kernelParameters = c(0.6))
lmodels = list( modelcont, modelcat)

model <- kmmMixedData(ldata, lmodels, nbCluster=2:5, strategy = clusterFastStrategy())
# plot only the first continuous data set
plot(model, y=c(1))

Plotting of a class [KmmModel]

Description

Plotting data from a [KmmModel] object using the estimated parameters and partition.

Usage

## S4 method for signature 'KmmModel'
plot(x, y, ...)

Arguments

See Also

plot

Examples

## the bull eyes data set
  data(bullsEye)
  model <- kmm( bullsEye, 2, models= "kmm_pk_s")
  plot(model)

Description

Print a MixAll S4 class to standard output.

Usage

## S4 method for signature 'ClusterAlgo'
print(x, ...)

## S4 method for signature 'ClusterAlgoPredict'
print(x, ...)

## S4 method for signature 'ClusterInit'
print(x, ...)

## S4 method for signature 'ClusterStrategy'
print(x, ...)

## S4 method for signature 'IClusterComponent'
print(x, ...)

## S4 method for signature 'IClusterModel'
print(x, ...)

## S4 method for signature 'ClusterCategoricalComponent'
print(x, k, ...)

## S4 method for signature 'ClusterCategorical'
print(x, ...)

## S4 method for signature 'ClusterDiagGaussianComponent'
print(x, k, ...)

## S4 method for signature 'ClusterDiagGaussian'
print(x, ...)

## S4 method for signature 'ClusterGammaComponent'
print(x, k, ...)

## S4 method for signature 'ClusterGamma'
print(x, ...)

## S4 method for signature 'ClusterMixedDataModel'
print(x, ...)

## S4 method for signature 'ClusterPoissonComponent'
print(x, k, ...)

## S4 method for signature 'ClusterPoisson'
print(x, ...)

## S4 method for signature 'IClusterPredict'
print(x, ...)

## S4 method for signature 'ClusterPredict'
print(x, ...)

## S4 method for signature 'ClusterPredictMixedData'
print(x, ...)

## S4 method for signature 'LearnAlgo'
print(x, ...)

## S4 method for signature 'KmmComponent'
print(x, k, ...)

## S4 method for signature 'KmmModel'
print(x, ...)

## S4 method for signature 'KmmMixedDataModel'
print(x, ...)

Arguments

Value

NULL. Prints to standard out.

See Also

print

Examples

## for cluster strategy
  strategy <- clusterStrategy()
  print(strategy)
  ## for cluster init
  init <- clusterInit()
  print(init)
  ## for cluster algo
  algo <- clusterAlgo()
  print(algo)

Description

Show description of a MixAll S4 class to standard output.

Usage

## S4 method for signature 'ClusterAlgo'
show(object)

## S4 method for signature 'ClusterAlgoPredict'
show(object)

## S4 method for signature 'ClusterInit'
show(object)

## S4 method for signature 'ClusterStrategy'
show(object)

## S4 method for signature 'IClusterComponent'
show(object)

## S4 method for signature 'IClusterModel'
show(object)

## S4 method for signature 'ClusterCategoricalComponent'
show(object)

## S4 method for signature 'ClusterCategorical'
show(object)

## S4 method for signature 'ClusterDiagGaussianComponent'
show(object)

## S4 method for signature 'ClusterDiagGaussian'
show(object)

## S4 method for signature 'ClusterGammaComponent'
show(object)

## S4 method for signature 'ClusterGamma'
show(object)

## S4 method for signature 'ClusterMixedDataModel'
show(object)

## S4 method for signature 'ClusterPoissonComponent'
show(object)

## S4 method for signature 'ClusterPoisson'
show(object)

## S4 method for signature 'IClusterPredict'
show(object)

## S4 method for signature 'ClusterPredict'
show(object)

## S4 method for signature 'ClusterPredictMixedData'
show(object)

## S4 method for signature 'LearnAlgo'
show(object)

## S4 method for signature 'KmmComponent'
show(object)

## S4 method for signature 'KmmModel'
show(object)

## S4 method for signature 'KmmMixedDataModel'
show(object)

Arguments

Value

NULL. Prints to standard out.

See Also

show

Examples

## for strategy
  strategy <- clusterStrategy()
  show(strategy)
  ## for cluster init
  init <- clusterInit()
  show(init)
  ## for cluster algo
  algo <- clusterAlgo()
  show(algo)

Description

Produce summary of a MixAll S4 class.

Usage

## S4 method for signature 'IClusterComponent'
summary(object, ...)

## S4 method for signature 'IClusterModel'
summary(object, ...)

## S4 method for signature 'ClusterCategoricalComponent'
summary(object)

## S4 method for signature 'ClusterCategorical'
summary(object, ...)

## S4 method for signature 'ClusterDiagGaussian'
summary(object, ...)

## S4 method for signature 'ClusterGamma'
summary(object, ...)

## S4 method for signature 'ClusterMixedDataModel'
summary(object, ...)

## S4 method for signature 'ClusterPoisson'
summary(object, ...)

## S4 method for signature 'IClusterPredict'
summary(object, ...)

## S4 method for signature 'ClusterPredict'
summary(object, ...)

## S4 method for signature 'ClusterPredictMixedData'
summary(object, ...)

## S4 method for signature 'KmmModel'
summary(object, ...)

## S4 method for signature 'KmmMixedDataModel'
summary(object, ...)

Arguments

Value

NULL. Summaries to standard out.