org.gephi.statistics.plugin.PageRank.java Source code

Java tutorial

Introduction

Here is the source code for org.gephi.statistics.plugin.PageRank.java

Source

/*
Copyright 2008-2011 Gephi
Authors : Patick J. McSweeney <pjmcswee@syr.edu>, Sebastien Heymann <seb@gephi.org>
Website : http://www.gephi.org
    
This file is part of Gephi.
    
DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS HEADER.
    
Copyright 2011 Gephi Consortium. All rights reserved.
    
The contents of this file are subject to the terms of either the GNU
General Public License Version 3 only ("GPL") or the Common
Development and Distribution License("CDDL") (collectively, the
"License"). You may not use this file except in compliance with the
License. You can obtain a copy of the License at
http://gephi.org/about/legal/license-notice/
or /cddl-1.0.txt and /gpl-3.0.txt. See the License for the
specific language governing permissions and limitations under the
License.  When distributing the software, include this License Header
Notice in each file and include the License files at
/cddl-1.0.txt and /gpl-3.0.txt. If applicable, add the following below the
License Header, with the fields enclosed by brackets [] replaced by
your own identifying information:
"Portions Copyrighted [year] [name of copyright owner]"
    
If you wish your version of this file to be governed by only the CDDL
or only the GPL Version 3, indicate your decision by adding
"[Contributor] elects to include this software in this distribution
under the [CDDL or GPL Version 3] license." If you do not indicate a
single choice of license, a recipient has the option to distribute
your version of this file under either the CDDL, the GPL Version 3 or
to extend the choice of license to its licensees as provided above.
However, if you add GPL Version 3 code and therefore, elected the GPL
Version 3 license, then the option applies only if the new code is
made subject to such option by the copyright holder.
    
Contributor(s):
    
Portions Copyrighted 2011 Gephi Consortium.
 */
package org.gephi.statistics.plugin;

import java.util.HashMap;
import java.util.Map;
import org.gephi.data.attributes.api.AttributeTable;
import org.gephi.data.attributes.api.AttributeColumn;
import org.gephi.data.attributes.api.AttributeModel;
import org.gephi.data.attributes.api.AttributeOrigin;
import org.gephi.data.attributes.api.AttributeRow;
import org.gephi.data.attributes.api.AttributeType;
import org.gephi.graph.api.Edge;
import org.gephi.graph.api.EdgeIterable;
import org.gephi.graph.api.GraphController;
import org.gephi.graph.api.GraphModel;
import org.gephi.graph.api.HierarchicalDirectedGraph;
import org.gephi.graph.api.HierarchicalGraph;
import org.gephi.graph.api.HierarchicalUndirectedGraph;
import org.gephi.graph.api.Node;
import org.gephi.statistics.spi.Statistics;
import org.gephi.utils.longtask.spi.LongTask;
import org.gephi.utils.progress.Progress;
import org.gephi.utils.progress.ProgressTicket;
import org.jfree.chart.ChartFactory;
import org.jfree.chart.JFreeChart;
import org.jfree.chart.plot.PlotOrientation;
import org.jfree.data.xy.XYSeries;
import org.jfree.data.xy.XYSeriesCollection;
import org.openide.util.Lookup;

/**
 * Ref: Sergey Brin, Lawrence Page, The Anatomy of a Large-Scale Hypertextual Web Search Engine, 
 * in Proceedings of the seventh International Conference on the World Wide Web (WWW1998):107-117
 *
 * @author pjmcswee
 */
public class PageRank implements Statistics, LongTask {

    public static final String PAGERANK = "pageranks";
    /** */
    private ProgressTicket progress;
    /** */
    private boolean isCanceled;
    /** */
    private double epsilon = 0.001;
    /** */
    private double probability = 0.85;
    private boolean useEdgeWeight = false;
    /** */
    private double[] pageranks;
    /** */
    private boolean isDirected;

    public PageRank() {
        GraphController graphController = Lookup.getDefault().lookup(GraphController.class);
        if (graphController != null && graphController.getModel() != null) {
            isDirected = graphController.getModel().isDirected();
        }
    }

    public void setDirected(boolean isDirected) {
        this.isDirected = isDirected;
    }

    /**
     *
     * @return
     */
    public boolean getDirected() {
        return isDirected;
    }

    public void execute(GraphModel graphModel, AttributeModel attributeModel) {
        HierarchicalGraph graph;
        if (isDirected) {
            graph = graphModel.getHierarchicalDirectedGraphVisible();
        } else {
            graph = graphModel.getHierarchicalUndirectedGraphVisible();
        }
        execute(graph, attributeModel);
    }

    public void execute(HierarchicalGraph hgraph, AttributeModel attributeModel) {
        isCanceled = false;

        hgraph.readLock();

        int N = hgraph.getNodeCount();
        pageranks = new double[N];
        double[] temp = new double[N];
        HashMap<Node, Integer> indicies = new HashMap<Node, Integer>();
        int index = 0;

        Progress.start(progress);
        double[] weights = null;
        if (useEdgeWeight) {
            weights = new double[N];
        }

        for (Node s : hgraph.getNodes()) {
            indicies.put(s, index);
            pageranks[index] = 1.0f / N;
            if (useEdgeWeight) {
                double sum = 0;
                EdgeIterable eIter;
                if (isDirected) {
                    eIter = ((HierarchicalDirectedGraph) hgraph).getOutEdgesAndMetaOutEdges(s);
                } else {
                    eIter = ((HierarchicalUndirectedGraph) hgraph).getEdgesAndMetaEdges(s);
                }
                for (Edge edge : eIter) {
                    sum += edge.getWeight();
                }
                weights[index] = sum;
            }
            index++;
        }

        while (true) {
            double r = 0;
            for (Node s : hgraph.getNodes()) {
                int s_index = indicies.get(s);
                boolean out;
                if (isDirected) {
                    out = ((HierarchicalDirectedGraph) hgraph).getTotalOutDegree(s) > 0;
                } else {
                    out = hgraph.getTotalDegree(s) > 0;
                }

                if (out) {
                    r += (1.0 - probability) * (pageranks[s_index] / N);
                } else {
                    r += (pageranks[s_index] / N);
                }
                if (isCanceled) {
                    hgraph.readUnlockAll();
                    return;
                }
            }

            boolean done = true;
            for (Node s : hgraph.getNodes()) {
                int s_index = indicies.get(s);
                temp[s_index] = r;

                EdgeIterable eIter;
                if (isDirected) {
                    eIter = ((HierarchicalDirectedGraph) hgraph).getInEdgesAndMetaInEdges(s);
                } else {
                    eIter = ((HierarchicalUndirectedGraph) hgraph).getEdgesAndMetaEdges(s);
                }

                for (Edge edge : eIter) {
                    Node neighbor = hgraph.getOpposite(s, edge);
                    int neigh_index = indicies.get(neighbor);
                    int normalize;
                    if (isDirected) {
                        normalize = ((HierarchicalDirectedGraph) hgraph).getTotalOutDegree(neighbor);
                    } else {
                        normalize = ((HierarchicalUndirectedGraph) hgraph).getTotalDegree(neighbor);
                    }
                    if (useEdgeWeight) {
                        double weight = edge.getWeight() / weights[neigh_index];
                        temp[s_index] += probability * pageranks[neigh_index] * weight;
                    } else {
                        temp[s_index] += probability * (pageranks[neigh_index] / normalize);
                    }

                }

                if ((temp[s_index] - pageranks[s_index]) / pageranks[s_index] >= epsilon) {
                    done = false;
                }

                if (isCanceled) {
                    hgraph.readUnlockAll();
                    return;
                }

            }
            pageranks = temp;
            temp = new double[N];
            if ((done) || (isCanceled)) {
                break;
            }

        }

        AttributeTable nodeTable = attributeModel.getNodeTable();
        AttributeColumn pangeRanksCol = nodeTable.getColumn(PAGERANK);
        if (pangeRanksCol == null) {
            pangeRanksCol = nodeTable.addColumn(PAGERANK, "PageRank", AttributeType.DOUBLE,
                    AttributeOrigin.COMPUTED, new Double(0));
        }

        for (Node s : hgraph.getNodes()) {
            int s_index = indicies.get(s);
            AttributeRow row = (AttributeRow) s.getNodeData().getAttributes();
            row.setValue(pangeRanksCol, pageranks[s_index]);
        }

        hgraph.readUnlockAll();
    }

    /**
     *
     * @return
     */
    public String getReport() {
        //distribution of values
        Map<Double, Integer> dist = new HashMap<Double, Integer>();
        for (int i = 0; i < pageranks.length; i++) {
            Double d = pageranks[i];
            if (dist.containsKey(d)) {
                Integer v = dist.get(d);
                dist.put(d, v + 1);
            } else {
                dist.put(d, 1);
            }
        }

        //Distribution series
        XYSeries dSeries = ChartUtils.createXYSeries(dist, "PageRanks");

        XYSeriesCollection dataset = new XYSeriesCollection();
        dataset.addSeries(dSeries);

        JFreeChart chart = ChartFactory.createXYLineChart("PageRank Distribution", "Score", "Count", dataset,
                PlotOrientation.VERTICAL, true, false, false);
        chart.removeLegend();
        ChartUtils.decorateChart(chart);
        ChartUtils.scaleChart(chart, dSeries, true);
        String imageFile = ChartUtils.renderChart(chart, "pageranks.png");

        String report = "<HTML> <BODY> <h1>PageRank Report </h1> " + "<hr> <br />" + "<h2> Parameters: </h2>"
                + "Epsilon = " + epsilon + "<br>" + "Probability = " + probability + "<br> <h2> Results: </h2>"
                + imageFile + "<br /><br />" + "<h2> Algorithm: </h2>"
                + "Sergey Brin, Lawrence Page, <i>The Anatomy of a Large-Scale Hypertextual Web Search Engine</i>, in Proceedings of the seventh International Conference on the World Wide Web (WWW1998):107-117<br />"
                + "</BODY> </HTML>";

        return report;

    }

    /**
     *
     * @return
     */
    public boolean cancel() {
        isCanceled = true;
        return true;
    }

    /**
     *
     * @param progressTicket
     */
    public void setProgressTicket(ProgressTicket progressTicket) {
        progress = progressTicket;
    }

    /**
     *
     * @param prob
     */
    public void setProbability(double prob) {
        probability = prob;
    }

    /**
     *
     * @param eps
     */
    public void setEpsilon(double eps) {
        epsilon = eps;
    }

    /**
     *
     * @return
     */
    public double getProbability() {
        return probability;
    }

    /**
     *
     * @return
     */
    public double getEpsilon() {
        return epsilon;
    }

    public boolean isUseEdgeWeight() {
        return useEdgeWeight;
    }

    public void setUseEdgeWeight(boolean useEdgeWeight) {
        this.useEdgeWeight = useEdgeWeight;
    }
}