Package elki.clustering.kmedoids

Class PAM<O>

  • Type Parameters:
    O - object datatype
    All Implemented Interfaces:
    elki.Algorithm, ClusteringAlgorithm<Clustering<MedoidModel>>, KMedoidsClustering<O>
    Direct Known Subclasses:
    CLARA, EagerPAM, FastPAM1, PAMSIL, ReynoldsPAM, SingleAssignmentKMedoids
    @Title("Partioning Around Medoids")
@Priority(100)
@Reference(authors="L. Kaufman, P. J. Rousseeuw",title="Clustering by means of Medoids",booktitle="Statistical Data Analysis Based on the L1-Norm and Related Methods",bibkey="books/misc/KauRou87") @Reference(authors="L. Kaufman, P. J. Rousseeuw",title="Partitioning Around Medoids (Program PAM)",booktitle="Finding Groups in Data: An Introduction to Cluster Analysis",url="https://doi.org/10.1002/9780470316801.ch2",bibkey="doi:10.1002/9780470316801.ch2") @Reference(authors="R. A. Whitaker",title="A Fast Algorithm For The Greedy Interchange For Large-Scale Clustering And Median Location Problems",booktitle="INFOR: Information Systems and Operational Research 21(2)",url="https://doi.org/10.1080/03155986.1983.11731889",bibkey="doi:10.1080/03155986.1983.11731889")
public class PAM<O>
extends java.lang.Object
implements KMedoidsClustering<O>
    The original Partitioning Around Medoids (PAM) algorithm or k-medoids clustering, as proposed by Kaufman and Rousseeuw; a largely equivalent method was also proposed by Whitaker in the operations research domain, and is well known by the name "fast interchange" there.

    Reference:

    L. Kaufman, P. J. Rousseeuw
    Clustering by means of Medoids
    Statistical Data Analysis Based on the L1-Norm and Related Methods

    R. A. Whitaker
    A Fast Algorithm For The Greedy Interchange For Large-Scale Clustering And Median Location Problems
    INFOR: Information Systems and Operational Research 21(2)

    Since:
    0.5.0
    Author:
    Erich Schubert

    • Nested Class Summary

      Nested Classes 
      Modifier and Type Class Description
      protected static class  PAM.Instance
      Instance for a single dataset.
      static class  PAM.Par<O>
      Parameterization class.

      • Nested classes/interfaces inherited from interface elki.Algorithm

        elki.Algorithm.Utils

    • Field Summary

      Fields 
      Modifier and Type Field Description
      protected elki.distance.Distance<? super O> distance
      Distance function used.
      protected KMedoidsInitialization<O> initializer
      Method to choose initial means.
      protected int k
      The number of clusters to produce.
      private static elki.logging.Logging LOG
      The logger for this class.
      protected int maxiter
      The maximum number of iterations.

    • Constructor Summary

      Constructors 
      Constructor Description
      PAM​(elki.distance.Distance<? super O> distance, int k, int maxiter, KMedoidsInitialization<O> initializer)
      Constructor.

    • Method Summary

      All Methods Static Methods Instance Methods Concrete Methods 
      Modifier and Type Method Description
      elki.data.type.TypeInformation[] getInputTypeRestriction()  
      protected elki.logging.Logging getLogger()
      Get the static class logger.
      protected elki.database.ids.ArrayModifiableDBIDs initialMedoids​(elki.database.query.distance.DistanceQuery<? super O> distQ, elki.database.ids.DBIDs ids, int k)
      Choose the initial medoids.
      Clustering<MedoidModel> run​(elki.database.relation.Relation<O> relation)
      Run k-medoids clustering.
      Clustering<MedoidModel> run​(elki.database.relation.Relation<O> relation, int k, elki.database.query.distance.DistanceQuery<? super O> distQ)
      Run k-medoids clustering with a given distance query.
      Not a very elegant API, but needed for some types of nested k-medoids.
      protected static Clustering<MedoidModel> wrapResult​(elki.database.ids.DBIDs ids, elki.database.datastore.WritableIntegerDataStore assignment, elki.database.ids.ArrayModifiableDBIDs medoids, java.lang.String name)
      Wrap the clustering result.

      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

    • Field Detail

      • LOG

        private static final elki.logging.Logging LOG
        The logger for this class.

      • distance

        protected elki.distance.Distance<? super O> distance
        Distance function used.

      • k

        protected int k
        The number of clusters to produce.

      • maxiter

        protected int maxiter
        The maximum number of iterations.

    • Constructor Detail

      • PAM

        public PAM​(elki.distance.Distance<? super O> distance,
           int k,
           int maxiter,
           KMedoidsInitialization<O> initializer)
        Constructor.
        Parameters:
        distance - distance function
        k - k parameter
        maxiter - Maxiter parameter
        initializer - Function to generate the initial means

    • Method Detail

      • run

        public Clustering<MedoidModel> run​(elki.database.relation.Relation<O> relation,
                                   int k,
                                   elki.database.query.distance.DistanceQuery<? super O> distQ)
        Description copied from interface: KMedoidsClustering
        Run k-medoids clustering with a given distance query.
        Not a very elegant API, but needed for some types of nested k-medoids.
        Specified by:
        run in interface KMedoidsClustering<O>
        Parameters:
        relation - relation to use
        k - Number of clusters
        distQ - Distance query to use
        Returns:
        result

      • initialMedoids

        protected elki.database.ids.ArrayModifiableDBIDs initialMedoids​(elki.database.query.distance.DistanceQuery<? super O> distQ,
                                                                elki.database.ids.DBIDs ids,
                                                                int k)
        Choose the initial medoids.
        Parameters:
        distQ - Distance query
        ids - IDs to choose from
        k - Number of medoids to choose from
        Returns:
        Initial medoids

      • wrapResult

        protected static Clustering<MedoidModel> wrapResult​(elki.database.ids.DBIDs ids,
                                                    elki.database.datastore.WritableIntegerDataStore assignment,
                                                    elki.database.ids.ArrayModifiableDBIDs medoids,
                                                    java.lang.String name)
        Wrap the clustering result.
        Parameters:
        ids - Object ids
        assignment - Cluster assignment
        medoids - Medoids
        name - Clustering name
        Returns:
        Wrapped result

      • getInputTypeRestriction

        public elki.data.type.TypeInformation[] getInputTypeRestriction()
        Specified by:
        getInputTypeRestriction in interface elki.Algorithm

      • getLogger

        protected elki.logging.Logging getLogger()
        Get the static class logger.