cluster_partitionAroundMedoids: A class to perform Partition Around Medoids clustering on...

In robertogattabs/pMiner.v045b: Processes Mining in Medicine

A class to perform Partition Around Medoids clustering on sequential data for Process Mining issues

Description

This class performs sequence clustering on an event-log with the Partition Around Medoids (PAM) algorithm. The public methods are:

• cluster_partitionAroundMedoids() is the constructor of the class

• loadDataset( ...) loads data taken from a dataLoader::getData() method, into a cluster_partitionAroundMedoids() object

• calculateClusters() performs the actual clustering computation on the previously loaded dataset

• getClusters() returns the clusters computed by the cluster_partitionAroundMedoids::calculateClusters() method

• getClusterStats( ... ) returns informations about the clustering result (i.e. support, between-cluster distance, within-cluster mean distance and standard deviation)

• getClusterLog( ... ) returns informations about the clustering computation itself (i.e. iterations needed to converge, centroids value after each iteration)

In order to better undestand the use of such methods, please visit: www.pminer.info

Parameters for cluster_partitionAroundMedoids::calculateClusters() method are:

• num the number of clusters it has to generate

• typeOfModel the name of the Process Mining model it has to use to generate the space (up to now, only the default "firstOrdermarkovModel" is provided)

Usage

cluster_partitionAroundMedoids()

Examples

## Not run: 

# create a Loader 
obj.L<-dataLoader();   

# Load a .csv using "DES" and "ID" as column names to indicate events
# and Patient's ID
obj.L$load.csv(nomeFile = "../otherFiles/test_02.csv",
IDName = "ID",EVENTName = "DES",dateColumnName = "DATA")

# now create an object cluster_partitionAroundMedoids
obj.clPAM<- cluster_partitionAroundMedoids();

# load the data into logInspector object
obj.clPAM$loadDataset( obj.L$getData() );

# perform clustering computation
obj.clPAM$calculateClusters(num = 2);

# get calculated clusters
a <- obj.clPAM$getClusters();

# get informations about performance of clusters
b <- obj.clPAM$getClusterStats();

# get log of each iteration of the algorithm
d <- obj.clPAM$getClusterLog();

## End(Not run)

robertogattabs/pMiner.v045b documentation built on Aug. 2, 2022, 1:53 p.m.