Matrix of simulated ordinal data

Description

This is a toy dataset for running simple examples.

Usage

Msimulated

Format

An ordinal data matrix with 60 lines and 50 columns. The number of levels is equal to 3. Four blocks are simulated with (mu,pi) parameters equal to (3,0.5), (2,0.7), (1,0.8) and (2,0.6).

Function to perform a classification

Description

This function performs a classification algorithm on a dataset with ordinal features, and a label variable that belongs to (1,2,...,kr). The classification function provides two classification models. The first model, (chosen by the argument kc=0), is a multivariate BOS model with the assumtion that, conditional on the class of the observations, the features are independent. The second model is a parsimonious version of the first model. Parsimony is introduced by grouping the features into clusters (as in co-clustering) and assuming that the features of a cluster have a common distribution.

Usage

bosclassif(x, y, idx_list=c(1), kr, kc=0, init, nbSEM, nbSEMburn, 
          nbindmini, m=0, percentRandomB=0) 

Arguments

Value

Return an object. The slots are:

Author(s)

Margot Selosse, Julien Jacques, Christophe Biernacki.

Examples

# loading the real dataset
data("dataqol.classif")

set.seed(5)

# loading the ordinal data
M <- as.matrix(dataqol.classif[,2:29])


# creating the classes values
y <- as.vector(dataqol.classif$death)


# sampling datasets for training and to predict
nb.sample <- ceiling(nrow(M)*2/3)
sample.train <- sample(1:nrow(M), nb.sample, replace=FALSE)

M.train <- M[sample.train,]
M.validation <- M[-sample.train,]
nb.missing.validation <- length(which(M.validation==0))
m <- c(4)
M.validation[which(M.validation==0)] <- sample(1:m, nb.missing.validation,replace=TRUE)


y.train <- y[sample.train]
y.validation <- y[-sample.train]



# configuration for SEM algorithm
nbSEM=50
nbSEMburn=40
nbindmini=1
init="kmeans"

# number of classes to predict
kr <- 2
# different kc to test with cross-validation
kcol <- 1


res <- bosclassif(x=M.train,y=y.train,kr=kr,kc=kcol,m=m,
                  nbSEM=nbSEM,nbSEMburn=nbSEMburn,
                  nbindmini=nbindmini,init=init)

predictions <- predict(res, M.validation)

Function to perform a clustering

Description

This function performs a clustering algorithm on ordinal data by using the multiple latent block model (see references for further details). It allows the user to define D groups of variables that have different numbers of levels. The BOS distribution is used, and the parameters inference is obtained using the SEM-Gibbs algorithm.

Usage

bosclust(x, idx_list=c(1), kr, init, nbSEM, nbSEMburn, 
        nbindmini, m=0, percentRandomB=0)

Arguments

Value

Author(s)

Margot Selosse, Julien Jacques, Christophe Biernacki.

Examples

  library(ordinalClust)
  data("dataqol")
  set.seed(5)

  # loading the ordinal data
  M <- as.matrix(dataqol[,2:29])

  m = 4

  krow = 4

  nbSEM=50
  nbSEMburn=40
  nbindmini=2
  init = "random"


  object <- bosclust(x=M,kr=krow, m=m, nbSEM=nbSEM,
      nbSEMburn=nbSEMburn, nbindmini=nbindmini, init=init)
    
  

Function to perform a co-clustering

Description

This function runs a co-clustering algorithm on ordinal data by using the latent block model (see references for further details). A BOS distribution is used, and the parameters inference is obtained using the SEM-Gbbs algorithm.

Usage

boscoclust(x=matrix(0,nrow=1,ncol=1), idx_list=c(1), kr, kc, init, nbSEM, nbSEMburn, 
          nbRepeat=1, nbindmini, m=0, percentRandomB=0)

Arguments

Value

Author(s)

Margot Selosse, Julien Jacques, Christophe Biernacki.

Examples

  
  
    
  library(ordinalClust)

  # loading the real dataset
  data("dataqol")
  set.seed(5)

  # loading the ordinal data
  M <- as.matrix(dataqol[,2:29])


  # defining different number of categories:
  m=4


  # defining number of row and column clusters
  krow = 4
  kcol = 4

  # configuration for the inference
  nbSEM=50
  nbSEMburn=40
  nbindmini=2
  init = "randomBurnin"
  percentRandomB=c(20,20)



  # Co-clustering execution
  object <- boscoclust(x=M,kr=krow,kc=kcol,m=m,nbSEM=nbSEM,
            nbSEMburn=nbSEMburn, nbindmini=nbindmini, init=init, percentRandomB=percentRandomB)

  
  

Questionnaire Responses Of Patients Affected By Breast Cancer

Description

This dataset contains the responses of 121 patients to 30 questions about their quality of life.

Usage

dataqol

Format

A dataframe with 121 lines and 31 columns. A line represents a patient and a column contains information about the patients.

Id

patient Id

q1-q28

responses to 28 questions with the number of levels equal to 4

q29-q30

responses to 22 questions with the number of levels equal to 7

Source

The table was determined based on data associated with the package available on: https://cran.r-project.org/package=QoLR

Questionnaire Responses Of Patients Affected By Breast Cancer

Description

This dataset contains the responses of 40 patients to 30 questions about their quality of life. Furthermore, a variable indicates if the patient survived the disease.

Usage

dataqol.classif

Format

A dataframe with 40 lines and 32 columns. A line represents a patient and a column contains information about the patients.

Id

patient Id

q1-q28

responses to 28 questions with the number of levels equal to 4

q29-q30

responses to 22 questions with the number of levels equal to 7

death

if the patient survived (1) or not (2)

Source

The table was determined based on data associated with the package available at: https://cran.r-project.org/package=QoLR

pejSim

Description

This function computes the probabilty for a level ej to be sampled from a BOS distribution of parameters (mu,pi), with the number of levels equal to m. It can be used to generate data with a BOS distribution.

Usage

pejSim(ej,m,mu,p)

Arguments

Value

Returns the probability of ej to be sampled from a BOS distribution of parameters (mu,pi), with the number of levels equal to m.

Author(s)

Margot Selosse, Julien Jacques, Christophe Biernacki.

Examples

  library(ordinalClust)
  data("dataqol")
  set.seed(5)

  m=7
  nr=10000
  mu=5
  pi=0.5

  probaBOS=rep(0,m)
  for (im in 1:m) probaBOS[im]=pejSim(im,m,mu,pi)
  M <- sample(1:m,nr,prob = probaBOS, replace=TRUE)

~~ Method for the function plot in the ordinalClust package ~~

Description

Plots the result of a classification, clustering or co-clustering performed using the following functions: bosclassif,bosclust,boscoclust.

Methods

signature(object = "ResultClassifOrdinal")

signature(object = "ResultClustOrdinal")

signature(object = "ResultCoclustOrdinal")

~~ Method for the function predict in the ordinalClust package ~~

Description

Method for the function predict in the ordinalClust package.

Methods

signature(object = "ResultClassifOrdinal")

Use this method with the result of the function bosclassif, and a new sample to predict the classes.

~~ Methods for the Function summary in the ordinalClust package ~~

Description

Prints the result of a classification, clustering or co-clustering performed using the following functions: bosclassif,bosclust,boscoclust.

Methods

signature(object = "ResultClassifOrdinal")

signature(object = "ResultClustOrdinal")

signature(object = "ResultCoclustOrdinal")