Example usage for org.apache.commons.math.stat.descriptive DescriptiveStatistics DescriptiveStatistics

List of usage examples for org.apache.commons.math.stat.descriptive DescriptiveStatistics DescriptiveStatistics

  1. HOME
  2. Java
  3. org.apache.commons
  4. org.apache.commons.math.stat.descriptive.*
  5. DescriptiveStatistics
  6. DescriptiveStatistics

Introduction

In this page you can find the example usage for org.apache.commons.math.stat.descriptive DescriptiveStatistics DescriptiveStatistics.

Prototype

public DescriptiveStatistics()

Source Link

Document

Construct a DescriptiveStatistics instance with an infinite window

Usage

From source file:org.matsim.contrib.socnetgen.sna.graph.analysis.TransitivityTask.java

/**
 * Calculates the mean local clustering coefficient, the maximum local
 * clustering coefficient, the minimum local clustering coefficient and the
 * global clustering coefficient. Writes the histogram of the local
 * clustering coefficient distribution into the output directory (if
 * specified).//from w w w. ja  v a2s. c om
 * 
 * @param graph
 *            a graph.
 * @param stats
 *            a map where the results of the analysis are stored.
 */
//   @Override
//   public void analyze(Graph graph, Map<String, Double> stats) {
//      DescriptiveStatistics distr = module.localClusteringDistribution(graph.getVertices());
//      double c_mean = distr.getMean();
//      double c_max = distr.getMax();
//      double c_min = distr.getMin();
//      stats.put(MEAN_LOCAL_CLUSTERING, c_mean);
//      stats.put(MAX_LOCAL_CLUSTERING, c_max);
//      stats.put(MIN_LOCAL_CLUSTERING, c_min);
//
//      double c_global = module.globalClusteringCoefficient(graph);
//      stats.put(GLOBAL_CLUSTERING_COEFFICIENT, c_global);
//
//      logger.info(String.format(
//            "c_local_mean = %1$.4f, c_local_max = %2$.4f, c_local_min = %3$.4f, c_global = %4$.4f.", c_mean, c_max,
//            c_min, c_global));
//
//      if (getOutputDirectory() != null) {
//         try {
//            writeHistograms(distr, new LinearDiscretizer(0.05), "c_local", false);
//         } catch (FileNotFoundException e) {
//            e.printStackTrace();
//         } catch (IOException e) {
//            e.printStackTrace();
//         }
//      }
//   }

public void analyze(Graph graph, Map<String, DescriptiveStatistics> statsMap) {
    String subKey = key + "_local";

    DescriptiveStatistics stats = module.statistics(graph.getVertices());
    printStats(stats, subKey);
    statsMap.put(subKey, stats);

    if (outputDirectoryNotNull()) {
        try {
            writeHistograms(stats, new LinearDiscretizer(0.05), subKey, false);
            writeHistograms(stats, subKey, 100, 20);
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    subKey = key + "_global";
    stats = new DescriptiveStatistics();
    stats.addValue(module.globalClusteringCoefficient(graph));
    printStats(stats, subKey);
    statsMap.put(subKey, stats);
}

From source file:org.matsim.contrib.socnetgen.sna.graph.matrix.MatrixAPL.java

public DescriptiveStatistics apl(AdjacencyMatrix<?> y) {
    List<APLThread> threads = new ArrayList<APLThread>();
    int size = (int) Math.floor(y.getVertexCount() / (double) numThreads);
    int i_start = 0;
    int i_stop = size;
    for (int i = 0; i < numThreads - 1; i++) {
        threads.add(//  w ww .  j av a 2 s .  c  o m
                new APLThread(y, new SingelPathDijkstra(y, new CostFunction()), i_start, i_stop, calcDistr));
        i_start = i_stop;
        i_stop += size;
    }
    threads.add(new APLThread(y, new SingelPathDijkstra(y, new CostFunction()), i_start, y.getVertexCount(),
            calcDistr));
    /*
     * start threads
     */
    logger.info(String.format("Calculating average path lenth on %1$s threads.", numThreads));
    ProgressLogger.init(y.getVertexCount(), 1, 5);
    for (APLThread thread : threads) {
        thread.start();
    }
    /*
     * wait for threads
     */
    for (APLThread thread : threads) {
        try {
            thread.join();
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
    }
    /*
     * merge results
     */
    DescriptiveStatistics stats = new DescriptiveStatistics();
    if (calcDistr) {
        for (APLThread thread : threads) {
            TDoubleArrayList vals = thread.getValues();
            for (int i = 0; i < vals.size(); i++) {
                stats.addValue(vals.get(i));
            }
        }
    } else {
        long lengthSum = 0;
        long pathCount = 0;
        for (APLThread thread : threads) {
            lengthSum += thread.lengthSum;
            pathCount += thread.pathCount;
        }
        stats.addValue(lengthSum / (double) pathCount);
    }

    return stats;
}

From source file:org.matsim.contrib.socnetgen.sna.graph.matrix.MatrixCentrality.java

/**
 * Builds the complete shortest path tree for all vertices and
 * simultaneously calculates closeness, betweenness, radius and diameter.
 * /*  w  w  w .j a  v  a  2 s .  c  o m*/
 * @param y an adjacency matrix.
 */
public void run(AdjacencyMatrix<?> y, int[] sources, int[] targets) {
    int n = y.getVertexCount();
    vertexCloseness = new double[n];
    Arrays.fill(vertexCloseness, Double.POSITIVE_INFINITY);
    vertexBetweenness = new double[n];
    edgeBetweenness = new TIntDoubleHashMap[n];
    diameter = 0;
    radius = Integer.MAX_VALUE;
    /*
     * create threads
     */
    if (dijkstraFactory == null)
        dijkstraFactory = new DijkstraFactory();

    List<CentralityThread> threads = new ArrayList<CentralityThread>();
    int size = (int) Math.floor(sources.length / (double) numThreads);
    int i_start = 0;
    int i_stop = size;
    for (int i = 0; i < numThreads - 1; i++) {
        int[] subSources = Arrays.copyOfRange(sources, i_start, i_stop);
        threads.add(new CentralityThread(y, subSources, targets, dijkstraFactory, calcBetweenness));
        i_start = i_stop;
        i_stop += size;
    }
    int[] subSources = Arrays.copyOfRange(sources, i_start, sources.length);
    threads.add(new CentralityThread(y, subSources, targets, dijkstraFactory, calcBetweenness));
    /*
     * start threads
     */
    for (CentralityThread thread : threads) {
        thread.start();
    }
    /*
     * wait for threads
     */
    for (CentralityThread thread : threads) {
        try {
            thread.join();
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
    }
    /*
     * merge results of threads
     */
    apl = new DescriptiveStatistics();
    double aplSum = 0;
    double cnt = 0;
    for (CentralityThread thread : threads) {

        if (calcAPLDistribution) {
            for (int i = 0; i < thread.pathLengths.size(); i++) {
                apl.addValue(thread.pathLengths.get(i));
            }
        } else {
            for (int i = 0; i < thread.pathLengths.size(); i++) {
                aplSum += thread.pathLengths.get(i);
            }
        }

        cnt++;

        for (int i = 0; i < n; i++) {
            /*
             * if this thread did not calculate the closeness of i it
             * returns infinity
             */
            vertexCloseness[i] = Math.min(vertexCloseness[i], thread.vertexCloseness[i]);
            /*
             * merge vertex betweenness values
             */
            vertexBetweenness[i] += thread.vertexBetweenness[i];
            /*
             * merge edge betweenness values
             */
            if (thread.edgeBetweenness[i] != null) {
                TIntDoubleIterator it = thread.edgeBetweenness[i].iterator();
                for (int j = 0; j < thread.edgeBetweenness[i].size(); j++) {
                    it.advance();
                    /*
                     * since we have an undirected graph, merge betweenness
                     * for both edges
                     */
                    if (edgeBetweenness[i] == null)
                        edgeBetweenness[i] = new TIntDoubleHashMap();
                    edgeBetweenness[i].adjustOrPutValue(it.key(), it.value(), it.value());

                    if (edgeBetweenness[it.key()] == null)
                        edgeBetweenness[it.key()] = new TIntDoubleHashMap();
                    edgeBetweenness[it.key()].adjustOrPutValue(i, it.value(), it.value());
                }
            }
            /*
             * get diameter and radius
             */
            diameter = Math.max(diameter, thread.diameter);
            radius = Math.min(radius, thread.radius);
        }
    }
    /*
     * calculate mean values
     */
    meanVertexCloseness = StatUtils.mean(vertexCloseness);

    if (!calcAPLDistribution)
        apl.addValue(aplSum / cnt);

    double sum = 0;
    for (int i = 0; i < y.getVertexCount(); i++)
        sum += vertexBetweenness[i];
    meanVertexBetweenness = sum / (double) y.getVertexCount();

    sum = 0;
    double count = 0;
    for (int i = 0; i < n; i++) {
        if (edgeBetweenness[i] != null) {
            TIntDoubleIterator it = edgeBetweenness[i].iterator();
            for (int k = 0; k < edgeBetweenness[i].size(); k++) {
                it.advance();
                sum += it.value();
                count++;
            }
        }
    }
    meanEdgeBetweenness = sum / count;
}

From source file:org.matsim.contrib.socnetgen.sna.graph.social.analysis.AgeTask.java

@SuppressWarnings("unchecked")
@Override/*from   w w  w.  ja  va2 s  .  c o  m*/
public void analyze(Graph graph, Map<String, DescriptiveStatistics> statsMap) {
    DescriptiveStatistics stats = module.statistics(graph.getVertices());
    statsMap.put(key, stats);
    printStats(stats, key);

    if (outputDirectoryNotNull()) {
        try {
            writeHistograms(stats, new LinearDiscretizer(1.0), key, false);

            StatsWriter.writeHistogram(module.correlation((Set<? extends SocialVertex>) graph.getVertices()),
                    "age", "age_mean", getOutputDirectory() + "/age_age.mean.txt");

            TDoubleObjectHashMap<DescriptiveStatistics> stat = module
                    .boxplot((Set<? extends SocialVertex>) graph.getVertices());
            StatsWriter.writeBoxplotStats(stat, getOutputDirectory() + "age_age.table.txt");
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    stats = new DescriptiveStatistics();
    stats.addValue(module.correlationCoefficient((Set<? extends SocialEdge>) graph.getEdges()));
    statsMap.put("r_" + key, stats);
    printStats(stats, "r_" + key);
}

From source file:org.matsim.contrib.socnetgen.sna.graph.social.analysis.GenderTask.java

@Override
public void analyze(Graph g, Map<String, DescriptiveStatistics> statsMap) {
    SocialGraph graph = (SocialGraph) g;

    Map<SocialVertex, String> values = module.values(graph.getVertices());
    TObjectDoubleHashMap<String> hist = LinguisticHistogram.create(values.values());

    DescriptiveStatistics male = new DescriptiveStatistics();
    male.addValue(hist.get(Gender.MALE));
    String key = "n_male";
    statsMap.put(key, male);//from   w  w  w .  j a v a2 s.  co  m
    printStats(male, key);

    DescriptiveStatistics female = new DescriptiveStatistics();
    female.addValue(hist.get(Gender.FEMALE));
    key = "n_female";
    statsMap.put(key, female);
    printStats(female, key);

    DescriptiveStatistics r = new DescriptiveStatistics();
    r.addValue(module.correlation(graph.getEdges()));
    key = "r_gender";
    statsMap.put(key, r);
    printStats(r, key);

    if (outputDirectoryNotNull()) {
        try {
            StatsWriter.writeLabeledHistogram(hist, "gender", "n",
                    String.format("%1$s/gender.txt", getOutputDirectory()));

            SocioMatrix<String> m = module.countsMatrix(graph.getVertices());
            m.toFile(String.format("%1$s/gender.countsMatrix.txt", getOutputDirectory()));

            SocioMatrixBuilder.normalizeTotalSum(m);
            m.toFile(String.format("%1$s/gender.countsMatrix.normTotal.txt", getOutputDirectory()));

            m = module.countsMatrix(graph.getVertices());
            SocioMatrixBuilder.normalizeRowSum(m);
            m.toFile(String.format("%1$s/gender.countsMatrix.normRow.txt", getOutputDirectory()));

            m = module.probaMatrix(graph.getVertices());
            m.toFile(String.format("%1$s/gender.probaMatrix.txt", getOutputDirectory()));
        } catch (IOException e) {
            e.printStackTrace();
        }
    }
}

From source file:org.matsim.contrib.socnetgen.sna.graph.spatial.analysis.Distance.java

public DescriptiveStatistics statistics(Set<? extends SpatialVertex> vertices) {
    DescriptiveStatistics stats = new DescriptiveStatistics();
    for (double val : distribution(vertices).getValues()) {
        stats.addValue(val);
    }/*from  w  w  w  . j a v  a 2 s  .  co  m*/
    return stats;
}

From source file:org.matsim.contrib.socnetgen.sna.graph.spatial.analysis.EdgeLengthMedian.java

private double edgeLengthMedian(SpatialVertex vertex) {
    if (calculator == null) {
        calculator = DistanceCalculatorFactory
                .createDistanceCalculator(CRSUtils.getCRS(vertex.getPoint().getSRID()));
    }/*  w  w w .j a  v a  2 s.  c  om*/

    DescriptiveStatistics stats = new DescriptiveStatistics();

    for (SpatialVertex neighbor : vertex.getNeighbours()) {
        if (neighbor.getPoint() != null) {
            if (vertex.getPoint().getSRID() == neighbor.getPoint().getSRID()) {
                stats.addValue(calculator.distance(vertex.getPoint(), neighbor.getPoint()));
            } else {
                throw new RuntimeException("Points do not share the same coordinate reference system.");
            }
        }
    }

    if (stats.getN() > 1)
        return stats.getPercentile(50);
    else
        return Double.NaN;
}

From source file:org.matsim.contrib.socnetgen.sna.snowball.analysis.BridgeEdgeTask.java

@Override
public void analyze(Graph g, Map<String, DescriptiveStatistics> results) {
    SampledGraph graph = (SampledGraph) g;

    int count = 0;

    for (SampledEdge e : graph.getEdges()) {
        SampledVertex v1 = e.getVertices().getFirst();
        SampledVertex v2 = e.getVertices().getSecond();

        if (v1.getSeed() != v2.getSeed())
            count++;/*  w w  w  .  j  ava 2  s  .  c  o m*/
    }

    DescriptiveStatistics stats = new DescriptiveStatistics();
    stats.addValue(count);

    printStats(stats, KEY);
    results.put(KEY, stats);
}

From source file:org.matsim.contrib.socnetgen.sna.snowball.analysis.EstimatorTask.java

@Override
public void analyze(Graph graph, Map<String, DescriptiveStatistics> stats) {
    TIntDoubleHashMap probas = new TIntDoubleHashMap(graph.getVertices().size());
    TIntIntHashMap counts = new TIntIntHashMap(graph.getVertices().size());

    double N_estim = 0;
    for (Vertex vertex : graph.getVertices()) {
        if (((SampledVertex) vertex).isSampled()) {
            int k = vertex.getNeighbours().size();
            double p = estimator.probability((SampledVertex) vertex);

            probas.adjustOrPutValue(k, p, p);
            counts.adjustOrPutValue(k, 1, 1);

            if (p > 0)
                N_estim += 1 / p;//from w ww. j  a v a  2s .  c  o m
        }
    }
    DescriptiveStatistics ds = new DescriptiveStatistics();
    ds.addValue(N_estim);
    stats.put("N_estim", ds);

    double M_estim = 0;
    for (Edge edge : graph.getEdges()) {
        SampledVertex v_i = (SampledVertex) edge.getVertices().getFirst();
        SampledVertex v_j = (SampledVertex) edge.getVertices().getSecond();
        if (v_i.isSampled() && v_j.isSampled()) {
            double p_i = estimator.probability(v_i);
            double p_j = estimator.probability(v_j);
            if (p_i > 0 && p_j > 0) {
                M_estim += 1 / ((p_i + p_j) - (p_i * p_j));
                //               M_estim += 1/(p_i * p_j);
            }
        }
    }
    ds = new DescriptiveStatistics();
    ds.addValue(M_estim);
    stats.put("M_estim", ds);

    TIntDoubleIterator it = probas.iterator();
    for (int i = 0; i < probas.size(); i++) {
        it.advance();
        it.setValue(it.value() / counts.get(it.key()));
    }

    writeValues(probas, "proba");
}

From source file:org.matsim.contrib.socnetgen.sna.snowball.analysis.ObservedTransitivity.java

/**
 * Returns a descriptive statistics object containing the local clustering
 * coefficient of all sampled vertices that have been sampled up to and
 * including the next to last iteration.
 * /*  w ww .j a  va 2s . com*/
 * @param vertices
 *            a set of sampled vertices
 * @return a descriptive statistics object.
 */
@SuppressWarnings("unchecked")
@Override
public DescriptiveStatistics statistics(Set<? extends Vertex> vertices) {
    DescriptiveStatistics stats = new DescriptiveStatistics();
    int iteration = SnowballStatistics.getInstance().lastIteration((Set<? extends SampledVertex>) vertices);

    TObjectDoubleHashMap<? extends Vertex> coefficients = localClusteringCoefficients(vertices);
    TObjectDoubleIterator<? extends Vertex> it = coefficients.iterator();

    for (int i = 0; i < coefficients.size(); i++) {
        it.advance();
        if (((SampledVertex) it.key()).getIterationSampled() <= iteration - 1) {
            stats.addValue(it.value());
        }
    }

    return stats;
}