Examples with LargeScaleSimulation edu.cmu.tetrad.sem.LargeScaleSimulation used on opensource projects

Example 1 with LargeScaleSimulation

use of edu.cmu.tetrad.sem.LargeScaleSimulation in project tetrad by cmu-phil.

the class TestGFci method test1.

public void test1() {
    RandomUtil.getInstance().setSeed(1450189593459L);
    int numNodes = 10;
    int numLatents = 5;
    int numEdges = 10;
    int sampleSize = 1000;
    double alpha = 0.01;
    double penaltyDiscount = 2;
    int depth = -1;
    int maxPathLength = -1;
    List<Node> vars = new ArrayList<>();
    for (int i = 0; i < numNodes; i++) {
        vars.add(new ContinuousVariable("X" + (i + 1)));
    }
    Graph dag = GraphUtils.randomGraphUniform(vars, numLatents, numEdges, 4, 4, 4, false);
    // Graph dag = GraphUtils.randomGraphRandomForwardEdges1(vars, numLatents, numEdges);
    // Graph dag = DataGraphUtils.scaleFreeGraph(vars, numLatents, .05, .05, .05, 3);
    DataSet data;
    LargeScaleSimulation simulator = new LargeScaleSimulation(dag);
    simulator.setCoefRange(.5, 1.5);
    simulator.setVarRange(1, 3);
    data = simulator.simulateDataFisher(sampleSize);
    data = DataUtils.restrictToMeasured(data);
    ICovarianceMatrix cov = new CovarianceMatrix(data);
    IndTestFisherZ independenceTest = new IndTestFisherZ(cov, alpha);
    SemBicScore score = new SemBicScore(cov);
    score.setPenaltyDiscount(penaltyDiscount);
    independenceTest.setAlpha(alpha);
    GFci gFci = new GFci(independenceTest, score);
    gFci.setVerbose(false);
    gFci.setMaxDegree(depth);
    gFci.setMaxPathLength(maxPathLength);
    gFci.setCompleteRuleSetUsed(false);
    gFci.setFaithfulnessAssumed(true);
    Graph outGraph = gFci.search();
    final DagToPag dagToPag = new DagToPag(dag);
    dagToPag.setCompleteRuleSetUsed(false);
    dagToPag.setMaxPathLength(maxPathLength);
    Graph truePag = dagToPag.convert();
    outGraph = GraphUtils.replaceNodes(outGraph, truePag.getNodes());
    int[][] counts = SearchGraphUtils.graphComparison(outGraph, truePag, null);
    int[][] expectedCounts = { { 0, 0, 0, 0, 0, 0 }, { 0, 4, 0, 0, 0, 1 }, { 0, 0, 3, 0, 0, 1 }, { 0, 0, 0, 0, 0, 0 }, { 0, 0, 0, 0, 0, 0 }, { 0, 0, 0, 0, 0, 0 }, { 0, 0, 0, 0, 0, 0 }, { 0, 0, 0, 0, 0, 0 } };
    for (int i = 0; i < counts.length; i++) {
        assertTrue(Arrays.equals(counts[i], expectedCounts[i]));
    }
// System.out.println(MatrixUtils.toString(counts));
// System.out.println(MatrixUtils.toString(expectedCounts));
}

Example 2 with LargeScaleSimulation

use of edu.cmu.tetrad.sem.LargeScaleSimulation in project tetrad by cmu-phil.

the class PerformanceTests method testPc.

public void testPc(int numVars, double edgeFactor, int numCases, double alpha) {
    int depth = -1;
    init(new File("long.pc." + numVars + "." + edgeFactor + "." + alpha + ".txt"), "Tests performance of the PC algorithm");
    long time1 = System.currentTimeMillis();
    System.out.println("Making list of vars");
    List<Node> vars = new ArrayList<>();
    for (int i = 0; i < numVars; i++) {
        vars.add(new ContinuousVariable("X" + i));
    }
    System.out.println("Finishing list of vars");
    System.out.println("Making graph");
    Graph graph = GraphUtils.randomGraphRandomForwardEdges(vars, 0, (int) (numVars * edgeFactor), 30, 15, 15, false, true);
    System.out.println("Graph done");
    out.println("Graph done");
    System.out.println("Starting simulation");
    LargeScaleSimulation simulator = new LargeScaleSimulation(graph);
    simulator.setOut(out);
    DataSet data = simulator.simulateDataFisher(numCases);
    System.out.println("Finishing simulation");
    long time2 = System.currentTimeMillis();
    out.println("Elapsed (simulating the data): " + (time2 - time1) + " ms");
    System.out.println("Making covariance matrix");
    // ICovarianceMatrix cov = new CovarianceMatrix2(data);
    ICovarianceMatrix cov = new CovarianceMatrixOnTheFly(data);
    // ICovarianceMatrix cov = new CorreqlationMatrix(new CovarianceMatrix2(data));
    // ICovarianceMatrix cov = new CovarianceMatrixOnTheFly(data, false);
    // ICovarianceMatrix cov = DataUtils.covarianceParanormalDrton(data);
    // ICovarianceMatrix cov = new CovarianceMatrix(DataUtils.covarianceParanormalWasserman(data));
    // System.out.println(cov);
    System.gc();
    System.out.println("Covariance matrix done");
    long time3 = System.currentTimeMillis();
    out.println("Elapsed (calculating cov): " + (time3 - time2) + " ms");
    // out.println(cov);
    IndTestFisherZ test = new IndTestFisherZ(cov, alpha);
    Pc pc = new Pc(test);
    pc.setVerbose(false);
    pc.setDepth(depth);
    // pcStable.setOut(out);
    Graph outGraph = pc.search();
    out.println(outGraph);
    long time4 = System.currentTimeMillis();
    out.println("# Vars = " + numVars);
    out.println("# Edges = " + (int) (numVars * edgeFactor));
    out.println("# Cases = " + numCases);
    out.println("Elapsed (simulating the data): " + (time2 - time1) + " ms");
    out.println("Elapsed (calculating cov): " + (time3 - time2) + " ms");
    out.println("Elapsed (running PC-Stable) " + (time4 - time3) + " ms");
    out.println("Total elapsed (cov + PC-Stable) " + (time4 - time2) + " ms");
    SearchGraphUtils.graphComparison(outGraph, SearchGraphUtils.patternForDag(graph), out);
    out.close();
}

Example 3 with LargeScaleSimulation

use of edu.cmu.tetrad.sem.LargeScaleSimulation in project tetrad by cmu-phil.

the class PerformanceTests method printStuffForKlea.

public void printStuffForKlea() {
    try {
        File _data = new File("data.txt");
        File _graph = new File("graph.txt");
        PrintStream out1 = new PrintStream(new FileOutputStream(_data));
        PrintStream out2 = new PrintStream(new FileOutputStream(_graph));
        int numVars = 50000;
        List<Node> vars = new ArrayList<>();
        for (int i = 0; i < numVars; i++) {
            vars.add(new ContinuousVariable("X" + (i + 1)));
        }
        double edgeFactor = 1.0;
        int numCases = 1000;
        Graph graph = GraphUtils.randomGraphRandomForwardEdges(vars, 0, (int) (numVars * edgeFactor), 30, 15, 15, false, true);
        out2.println(graph);
        System.out.println("Graph done");
        out.println("Graph done");
        System.out.println("Starting simulation");
        LargeScaleSimulation simulator = new LargeScaleSimulation(graph);
        simulator.setOut(out);
        DataSet data = simulator.simulateDataFisher(numCases);
        out1.println(data);
        out1.close();
        out2.close();
    } catch (Exception e) {
    }
}

Example 4 with LargeScaleSimulation

use of edu.cmu.tetrad.sem.LargeScaleSimulation in project tetrad by cmu-phil.

the class PerformanceTests method testCpc.

public void testCpc(int numVars, double edgeFactor, int numCases) {
    double alpha = 0.0001;
    int depth = -1;
    init(new File("long.cpc." + numVars + ".txt"), "Tests performance of the CPC algorithm");
    long time1 = System.currentTimeMillis();
    System.out.println("Making list of vars");
    List<Node> vars = new ArrayList<>();
    for (int i = 0; i < numVars; i++) {
        vars.add(new ContinuousVariable("X" + i));
    }
    System.out.println("Finishing list of vars");
    System.out.println("Making graph");
    Graph graph = GraphUtils.randomGraphRandomForwardEdges(vars, 0, (int) (numVars * edgeFactor), 30, 15, 15, false, true);
    System.out.println("Graph done");
    out.println("Graph done");
    System.out.println("Starting simulation");
    LargeScaleSimulation simulator = new LargeScaleSimulation(graph);
    simulator.setOut(out);
    DataSet data = simulator.simulateDataFisher(numCases);
    System.out.println("Finishing simulation");
    long time2 = System.currentTimeMillis();
    out.println("Elapsed (simulating the data): " + (time2 - time1) + " ms");
    System.out.println("Making covariance matrix");
    // ICovarianceMatrix cov = new CovarianceMatrix2(data);
    ICovarianceMatrix cov = new CovarianceMatrixOnTheFly(data);
    // ICovarianceMatrix cov = new CorreqlationMatrix(new CovarianceMatrix2(data));
    // ICovarianceMatrix cov = new CovarianceMatrixOnTheFly(data, false);
    // ICovarianceMatrix cov = DataUtils.covarianceParanormalDrton(data);
    // ICovarianceMatrix cov = new CovarianceMatrix(DataUtils.covarianceParanormalWasserman(data));
    // System.out.println(cov);
    System.gc();
    System.out.println("Covariance matrix done");
    long time3 = System.currentTimeMillis();
    out.println("Elapsed (calculating cov): " + (time3 - time2) + " ms");
    // out.println(cov);
    IndTestFisherZ test = new IndTestFisherZ(cov, alpha);
    Cpc cpc = new Cpc(test);
    cpc.setVerbose(false);
    cpc.setDepth(depth);
    // pcStable.setOut(out);
    Graph outGraph = cpc.search();
    out.println(outGraph);
    long time4 = System.currentTimeMillis();
    out.println("# Vars = " + numVars);
    out.println("# Edges = " + (int) (numVars * edgeFactor));
    out.println("# Cases = " + numCases);
    out.println("Elapsed (simulating the data): " + (time2 - time1) + " ms");
    out.println("Elapsed (calculating cov): " + (time3 - time2) + " ms");
    out.println("Elapsed (running PC-Stable) " + (time4 - time3) + " ms");
    out.println("Total elapsed (cov + PC-Stable) " + (time4 - time2) + " ms");
    SearchGraphUtils.graphComparison(outGraph, SearchGraphUtils.patternForDag(graph), out);
    out.close();
}

Example 5 with LargeScaleSimulation

use of edu.cmu.tetrad.sem.LargeScaleSimulation in project tetrad by cmu-phil.

the class PerformanceTests method testGFciComparison.

public void testGFciComparison(int numVars, double edgeFactor, int numCases, int numLatents) {
    numVars = 1000;
    edgeFactor = 1.0;
    numLatents = 100;
    numCases = 1000;
    int numRuns = 5;
    double alpha = 0.01;
    double penaltyDiscount = 3.0;
    int depth = 3;
    int maxPathLength = 3;
    boolean possibleDsepDone = true;
    boolean completeRuleSetUsed = false;
    boolean faithfulnessAssumed = true;
    init(new File("fci.algorithm.comparison" + numVars + "." + (int) (edgeFactor * numVars) + "." + numCases + ".txt"), "Num runs = " + numRuns);
    out.println("Num vars = " + numVars);
    out.println("Num edges = " + (int) (numVars * edgeFactor));
    out.println("Num cases = " + numCases);
    out.println("Alpha = " + alpha);
    out.println("Penalty discount = " + penaltyDiscount);
    out.println("Depth = " + depth);
    out.println("Maximum reachable path length for dsep search and discriminating undirectedPaths = " + maxPathLength);
    out.println("Num additional latent common causes = " + numLatents);
    out.println("Possible Dsep Done = " + possibleDsepDone);
    out.println("Complete Rule Set Used = " + completeRuleSetUsed);
    out.println();
    List<GraphUtils.GraphComparison> ffciCounts = new ArrayList<>();
    List<double[]> ffciArrowStats = new ArrayList<>();
    List<double[]> ffciTailStats = new ArrayList<>();
    List<Long> ffciElapsedTimes = new ArrayList<>();
    for (int run = 0; run < numRuns; run++) {
        out.println("\n\n\n******************************** RUN " + (run + 1) + " ********************************\n\n");
        System.out.println("Making list of vars");
        List<Node> vars = new ArrayList<>();
        for (int i = 0; i < numVars; i++) {
            vars.add(new ContinuousVariable("X" + (i + 1)));
        }
        System.out.println("Finishing list of vars");
        Graph dag = getLatentGraph(vars, edgeFactor, numLatents);
        System.out.println("Graph done");
        final DagToPag dagToPag = new DagToPag(dag);
        dagToPag.setCompleteRuleSetUsed(false);
        dagToPag.setMaxPathLength(maxPathLength);
        Graph truePag = dagToPag.convert();
        System.out.println("True PAG_of_the_true_DAG done");
        // Data.
        System.out.println("Starting simulation");
        LargeScaleSimulation simulator = new LargeScaleSimulation(dag);
        simulator.setCoefRange(.5, 1.5);
        simulator.setVarRange(1, 3);
        DataSet data = simulator.simulateDataFisher(numCases);
        data = DataUtils.restrictToMeasured(data);
        System.out.println("Finishing simulation");
        System.out.println("Making covariance matrix");
        ICovarianceMatrix cov = new CovarianceMatrixOnTheFly(data);
        System.out.println("Covariance matrix done");
        // Independence test.
        final IndTestFisherZ independenceTest = new IndTestFisherZ(cov, alpha);
        final SemBicScore score = new SemBicScore(cov);
        score.setPenaltyDiscount(penaltyDiscount);
        Graph estPag;
        long elapsed;
        // out.println("\n\n\n========================FCI run " + (run + 1));
        out.println("\n\n\n========================TGFCI run " + (run + 1));
        long ta1 = System.currentTimeMillis();
        // FCI fci = new FCI(independenceTest);
        GFci fci = new GFci(independenceTest, score);
        // TFci fci = new TFci(independenceTest);
        // fci.setVerbose(false);
        fci.setMaxDegree(depth);
        fci.setMaxPathLength(maxPathLength);
        // fci.setPossibleDsepSearchDone(possibleDsepDone);
        fci.setCompleteRuleSetUsed(completeRuleSetUsed);
        fci.setFaithfulnessAssumed(faithfulnessAssumed);
        estPag = fci.search();
        long ta2 = System.currentTimeMillis();
        estPag = GraphUtils.replaceNodes(estPag, truePag.getNodes());
        Set<Node> missingNodes = new HashSet<>();
        for (Node node : dag.getNodes()) {
            if (!estPag.containsNode(node)) {
                missingNodes.add(node);
            }
        }
        ffciArrowStats.add(printCorrectArrows(dag, estPag, truePag));
        ffciTailStats.add(printCorrectTails(dag, estPag, truePag));
        ffciCounts.add(SearchGraphUtils.getGraphComparison2(estPag, truePag));
        elapsed = ta2 - ta1;
        ffciElapsedTimes.add(elapsed);
        out.println("\nElapsed: " + elapsed + " ms");
        try {
            PrintStream out2 = new PrintStream(new File("dag." + run + ".txt"));
            out2.println(dag);
            PrintStream out3 = new PrintStream(new File("estpag." + run + ".txt"));
            out3.println(estPag);
            PrintStream out4 = new PrintStream(new File("truepag." + run + ".txt"));
            out4.println(truePag);
            out2.close();
            out3.close();
            out4.close();
        } catch (FileNotFoundException e) {
            e.printStackTrace();
            throw new RuntimeException(e);
        }
    }
    // printAverageConfusion("Average", ffciCounts);
    printAverageStatistics(ffciElapsedTimes, new ArrayList<Double>());
    out.close();
}