clusterAlgorithmPam-class {goCluster}R Documentation

Class "clusterAlgorithmPam", provides partitioning around medoids for goCluster.

Description

This can be used to group a dataset according to partitioning around medoids. The resulting gene groups can subsequently be analysed for significant enrichment of specific annotations.

Details

The class provides a wrapper around the clusterpam function. Please read the corresponding documentation for further details.

Slots

clusters:
Object of class "numeric", determines the number of clusters partitioning around medoids will identify.
distance:
Object of class "character", specifies the distance matrix that will be used.

Additional slots are described in the documentation of the clusterAlgorithm-class and clusterModule-class.

Extends

Class "clusterAlgorithm", directly. Class "clusterModule", by class "clusterAlgorithm".

Methods

config
signature(object = "clusterAlgorithmPam"): interactive setup of the class. You will be asked to specify the number of clusters PAM clustering should result in. In addition the distance matrix needs to be defined.
setup
signature(object = "clusterAlgorithmPam"): returns the configuration of the object as a list. This list can again be used for the non-interactive setup of the class.
setup<-
signature(object = "clusterAlgorithmPam"): non-interactive setup of the class. The options are specified using a list.
execute
signature(object = "clusterAlgorithmPam"): run the clustering.
reset
signature(object = "clusterAlgorithmPam"): remove all cluster data so that the execute function can be run again.
print
signature(object = "clusterAlgorithmPam"): This function prints some basic information about the content of this object.

Author(s)

Gunnar Wrobel, http://www.gunnarwrobel.de.

See Also

clusterpam, goCluster-class, clusterModule-class, clusterAlgorithm-class, clusterAlgorithmKmeans-class, clusterAlgorithmClara-class, clusterAlgorithmHclust-class, .

Examples

## Predefined setup for goCluster
## (This configurations selects the
## clara clustering)
data(benomylsetup)

## Change the setup to
## pam clustering
benomylsetup$data$dataset  <- benomylsetup$data$dataset[1:200,]
benomylsetup$data$uniqueid <- benomylsetup$data$uniqueid[1:200]
benomylsetup$classalgo <- "clusterAlgorithmPam"

## Setup a new goCluster object
test <- new("goCluster")
setup(test) <- benomylsetup

## Retrieve annotation
test@data <- execute(test@data, test)

## Cluster the dataset
test@algo <- execute(test@algo, test)

[Package goCluster version 1.0.3 Index]