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

Example 61 with SemPm

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

the class TestSemXml method sampleSemIm1.

private static SemIm sampleSemIm1() {
    List<Node> nodes = new ArrayList<>();
    for (int i = 0; i < 5; i++) {
        nodes.add(new ContinuousVariable("X" + (i + 1)));
    }
    Graph graph = new Dag(GraphUtils.randomGraph(nodes, 0, 5, 30, 15, 15, true));
    SemPm pm = new SemPm(graph);
    return new SemIm(pm);
}

Example 62 with SemPm

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

the class TestStatUtils method testConditionalCorrelation.

/**
 * Tests that the unconditional correlations and covariances are correct,
 * at least for the unconditional tests.
 */
@Test
public void testConditionalCorrelation() {
    RandomUtil.getInstance().setSeed(30299533L);
    // Make sure the unconditional correlations and covariances are OK.
    List<Node> nodes1 = new ArrayList<>();
    for (int i = 0; i < 5; i++) {
        nodes1.add(new ContinuousVariable("X" + (i + 1)));
    }
    Graph graph = new Dag(GraphUtils.randomGraph(nodes1, 0, 5, 3, 3, 3, false));
    SemPm pm = new SemPm(graph);
    SemIm im = new SemIm(pm);
    DataSet dataSet = im.simulateData(1000, false);
    double[] x = dataSet.getDoubleData().getColumn(0).toArray();
    double[] y = dataSet.getDoubleData().getColumn(1).toArray();
    double r1 = StatUtils.correlation(x, y);
    double s1 = StatUtils.covariance(x, y);
    double v1 = StatUtils.variance(x);
    double sd1 = StatUtils.sd(x);
    ICovarianceMatrix cov = new CovarianceMatrix(dataSet);
    TetradMatrix _cov = cov.getMatrix();
    double r2 = StatUtils.partialCorrelation(_cov, 0, 1);
    double s2 = StatUtils.partialCovariance(_cov, 0, 1);
    double v2 = StatUtils.partialVariance(_cov, 0);
    double sd2 = StatUtils.partialStandardDeviation(_cov, 0);
    assertEquals(r1, r2, .1);
    assertEquals(s1, s2, .1);
    assertEquals(v1, v2, .1);
    assertEquals(sd1, sd2, 0.1);
}

Example 63 with SemPm

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

the class TestDeltaSextadTest method test2.

@Test
public void test2() {
    int c = 2;
    int m = 2;
    int p = 6;
    Graph g = new EdgeListGraph();
    List<List<Node>> varClusters = new ArrayList<>();
    List<List<Node>> latents = new ArrayList<>();
    List<Node> vars = new ArrayList<>();
    for (int y = 0; y < c; y++) {
        varClusters.add(new ArrayList<Node>());
        latents.add(new ArrayList<Node>());
    }
    int e = 0;
    for (int y = 0; y < c; y++) {
        for (int i = 0; i < p; i++) {
            GraphNode n = new GraphNode("V" + ++e);
            vars.add(n);
            varClusters.get(y).add(n);
            g.addNode(n);
        }
    }
    List<Node> l = new ArrayList<>();
    int f = 0;
    for (int y = 0; y < c; y++) {
        for (int j = 0; j < m; j++) {
            Node _l = new GraphNode("L" + ++f);
            _l.setNodeType(NodeType.LATENT);
            l.add(_l);
            latents.get(y).add(_l);
            g.addNode(_l);
        }
    }
    for (int y = 0; y < c; y++) {
        for (int i = 0; i < m; i++) {
            for (int j = 0; j < p; j++) {
                g.addDirectedEdge(latents.get(y).get(i), varClusters.get(y).get(j));
            }
        }
    }
    for (int y = 1; y < c; y++) {
        for (int j = 0; j < m; j++) {
            g.addDirectedEdge(latents.get(y - 1).get(j), latents.get(y).get(j));
        }
    }
    SemPm pm = new SemPm(g);
    SemIm im = new SemIm(pm);
    DataSet data = im.simulateData(1000, false);
    List<Integer> indices = new ArrayList<>();
    indices.add(0);
    indices.add(1);
    indices.add(2);
    indices.add(4);
    indices.add(5);
    indices.add(7);
    Collections.shuffle(indices);
    // Node x1 = data.getVariable(indices.get(0));
    // Node x2 = data.getVariable(indices.get(1));
    // Node x3 = data.getVariable(indices.get(2));
    // Node x4 = data.getVariable(indices.get(3));
    // Node x5 = data.getVariable(indices.get(4));
    // Node x6 = data.getVariable(indices.get(5));
    int x1 = indices.get(0);
    int x2 = indices.get(1);
    int x3 = indices.get(2);
    int x4 = indices.get(3);
    int x5 = indices.get(4);
    int x6 = indices.get(5);
    DeltaSextadTest test = new DeltaSextadTest(data);
    // Should be invariant to changes or order of the first three or of the last three variables.
    double a = test.getPValue(new IntSextad(x1, x2, x3, x4, x5, x6));
    double b = test.getPValue(new IntSextad(x2, x3, x1, x5, x4, x6));
    assertEquals(a, b, 1e-7);
}

Example 64 with SemPm

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

the class TestDeltaSextadTest method getSem1.

private SemIm getSem1() {
    Graph graph = new EdgeListGraph();
    Node l1 = new GraphNode("l1");
    Node l2 = new GraphNode("l2");
    l1.setNodeType(NodeType.LATENT);
    l2.setNodeType(NodeType.LATENT);
    List<Node> measures = new ArrayList<>();
    int numMeasures = 8;
    for (int i = 0; i < numMeasures; i++) {
        measures.add(new GraphNode("X" + (i + 1)));
    }
    graph.addNode(l1);
    graph.addNode(l2);
    for (int i = 0; i < numMeasures; i++) {
        graph.addNode(measures.get(i));
        graph.addDirectedEdge(l1, measures.get(i));
        graph.addDirectedEdge(l2, measures.get(i));
    }
    SemPm pm = new SemPm(graph);
    Parameters params = new Parameters();
    return new SemIm(pm, params);
}

Example 65 with SemPm

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

the class TestPc method printStatsPcRegression.

private double[] printStatsPcRegression(String[] algorithms, int t, boolean directed, int numRuns, double alpha, double penaltyDiscount, int numMeasures, int numLatents, double edgeFactor, int sampleSize) {
    NumberFormat nf = new DecimalFormat("0.00");
    double sumAdjPrecision = 0.0;
    double sumAdjRecall = 0.0;
    int count = 0;
    for (int i = 0; i < numRuns; i++) {
        int numEdges = (int) (edgeFactor * (numMeasures + numLatents));
        List<Node> nodes = new ArrayList<>();
        List<String> names = new ArrayList<>();
        for (int r = 0; r < numMeasures + numLatents; r++) {
            String name = "X" + (r + 1);
            nodes.add(new ContinuousVariable(name));
            names.add(name);
        }
        Graph dag = GraphUtils.randomGraphRandomForwardEdges(nodes, numLatents, numEdges, 10, 10, 10, false);
        SemPm pm = new SemPm(dag);
        SemIm im = new SemIm(pm);
        DataSet data = im.simulateData(sampleSize, false);
        // Graph comparison = dag;
        Graph comparison = new DagToPag(dag).convert();
        // Graph comparison = new Pc(new IndTestDSep(dag)).search();
        IndTestFisherZ test = new IndTestFisherZ(data, alpha);
        SemBicScore score = new SemBicScore(new CovarianceMatrixOnTheFly(data));
        score.setPenaltyDiscount(penaltyDiscount);
        GraphSearch search;
        Graph out;
        Node target = null;
        for (Node node : nodes) {
            if (node.getNodeType() == NodeType.MEASURED) {
                target = node;
                break;
            }
        }
        switch(t) {
            case 0:
                search = new Pc(test);
                out = search.search();
                break;
            case 1:
                search = new Cpc(test);
                out = search.search();
                break;
            case 2:
                search = new Fges(score);
                out = search.search();
                break;
            case 3:
                search = new Fci(test);
                out = search.search();
                break;
            case 4:
                search = new GFci(test, score);
                out = search.search();
                break;
            case 5:
                search = new Rfci(test);
                out = search.search();
                break;
            case 6:
                search = new Cfci(test);
                out = search.search();
                break;
            case 7:
                out = getRegressionGraph(data, target);
                break;
            default:
                throw new IllegalStateException();
        }
        target = out.getNode(target.getName());
        out = trim(out, target);
        long start = System.currentTimeMillis();
        long stop = System.currentTimeMillis();
        long elapsed = stop - start;
        out = GraphUtils.replaceNodes(out, dag.getNodes());
        for (Node node : dag.getNodes()) {
            if (!out.containsNode(node)) {
                out.addNode(node);
            }
        }
        int adjTp = 0;
        int adjFp = 0;
        int adjFn = 0;
        for (Node node : out.getAdjacentNodes(target)) {
            if (comparison.isAdjacentTo(target, node)) {
                adjTp++;
            } else {
                adjFp++;
            }
        }
        for (Node node : dag.getAdjacentNodes(target)) {
            if (!out.isAdjacentTo(target, node)) {
                adjFn++;
            }
        }
        double adjPrecision = adjTp / (double) (adjTp + adjFp);
        double adjRecall = adjTp / (double) (adjTp + adjFn);
        if (!Double.isNaN(adjPrecision)) {
            sumAdjPrecision += adjPrecision;
        }
        if (!Double.isNaN(adjRecall)) {
            sumAdjRecall += adjRecall;
        }
        count++;
    }
    double avgAdjPrecision = sumAdjPrecision / (double) count;
    double avgAdjRecall = sumAdjRecall / (double) count;
    double[] ret = new double[] { avgAdjPrecision, avgAdjRecall };
    System.out.println();
    System.out.println(algorithms[t] + " adj precision " + nf.format(avgAdjPrecision));
    System.out.println(algorithms[t] + " adj recall " + nf.format(avgAdjRecall));
    return ret;
}