Examples of stallone.api.cluster.IClustering

stallone.api.cluster.IClustering
Each clustering algorithm implements this interface. It provides methods - to set input as an indexes set of vectors e.g. data points - to set clustering metric - to perform clustering - to obtain cluster assignment (either crips via getCluster or fuzzy via getMembership) of passed data points. @author Martin Senne

        names.add("/Users/noe/data/open_projects/adaptive_sampling_local/data/TrypsinBenzamidine/new_model_test/model/tics/long2.dat");
        names.add("/Users/noe/data/open_projects/adaptive_sampling_local/data/TrypsinBenzamidine/new_model_test/model/tics/long3.dat");
        IDataSequenceLoader loader = dataNew.multiSequenceLoader(names);
        loader.scan();
        
        IClustering clustering = clusterNew.regspace(loader, 0.5);
        clustering.perform();
        
        /*
        IDiscretization assign = clustering.getClusterAssignment();


        System.out.println("number of clusters: "+clustering.getNumberOfClusters());

View Full Code Here


        System.out.println("Data concatenated to size "+obscat.size());


        System.out.println("Clustering...");


        IClustering cluster = Cluster.util.densityBased(obscat, nstates);


        System.out.println("done.");


        IIntArray ci = cluster.getClusterIndexes();
        IDoubleArray[] res = new IDoubleArray[nstates];
        for (int state=0; state<nstates; state++)
        {
            IParameterEstimator estimator = Statistics.create.parameterEstimatorGaussian1D();
            res[state] = estimator.estimate(obscat, Cluster.util.membershipToState(cluster, state));

View Full Code Here

        {
            System.out.println(getUsageString());
            System.exit(0);
        }


        IClustering clustering1 = clusterNew.kmeans(cmd.nclusters, 10);
        IClustering clustering2 = clusterNew.kmeans(cmd.nsplit, 10);


        // split only on first trajectory:
        cmd.mc = new MultiClustering(cmd.data.getSequence(0), clustering1, clustering2);
        cmd.mc.split(0);
        cmd.mc.split(0);

View Full Code Here


        // create Multi-Level-Clustering 
        if (cmd.ninitclusters > 0)
        {
            // k-means initial clustering and random splitting
            IClustering clustering1 = clusterNew.kmeans(cmd.ninitclusters, 10);
            clustering1.setMetric(cmd.metric);
            IClustering clustering2 = clusterNew.random(cmd.nsplit);
            //IClustering clustering2 = clusterNew.createKcenter(cmd.nsplit);
            clustering2.setMetric(cmd.metric);
            // construct multiclustering
            //cmd.mc = new MultiClustering(cmd.data.get(0), clustering1, clustering2);
            cmd.mc = new MultiClusteringSplitMerge(cmd.data.getSequence(0), clustering1, clustering2);
        }
        else
        {
            IClustering clustering2 = clusterNew.randomCompact(cmd.nsplit, 10);
            //IClustering clustering2 = clusterNew.createRandom(cmd.nsplit);
            //IClustering clustering2 = clusterNew.createKcenter(cmd.nsplit);
            clustering2.setMetric(cmd.metric);
            cmd.mc = new MultiClusteringSplitMerge(cmd.data.getSequence(0), cmd.initcenters, cmd.metric, clustering2);
        }


        // fixed initial clustering
        /*

View Full Code Here

TOP

Related Classes of stallone.api.cluster.IClustering

stallone.api.hmm.HMMFactory

stallone.hmm.pmm.AdaptiveDiscretization

stallone.hmm.pmm.MultiClusteringKmeans

stallone.ui.Test

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.