Examples with LinearFisherModel edu.cmu.tetrad.algcomparison.simulation.LinearFisherModel used on opensource projects

Example 1 with LinearFisherModel

use of edu.cmu.tetrad.algcomparison.simulation.LinearFisherModel in project tetrad by cmu-phil.

the class ExampleFirstInflection method main.

public static void main(String... args) {
    Parameters parameters = new Parameters();
    parameters.set("numMeasures", 40, 100);
    parameters.set("avgDegree", 2);
    parameters.set("sampleSize", 400, 800);
    parameters.set("numRuns", 10);
    parameters.set("differentGraphs", true);
    parameters.set("numLatents", 0);
    parameters.set("maxDegree", 100);
    parameters.set("maxIndegree", 100);
    parameters.set("maxOutdegree", 100);
    parameters.set("connected", false);
    parameters.set("coefLow", 0.2);
    parameters.set("coefHigh", 0.9);
    parameters.set("varLow", 1);
    parameters.set("varHigh", 3);
    parameters.set("verbose", false);
    parameters.set("coefSymmetric", true);
    parameters.set("percentDiscrete", 0);
    parameters.set("numCategories", 3);
    parameters.set("differentGraphs", true);
    parameters.set("intervalBetweenShocks", 10);
    parameters.set("intervalBetweenRecordings", 10);
    parameters.set("fisherEpsilon", 0.001);
    parameters.set("randomizeColumns", true);
    parameters.set("alpha", 1e-8);
    parameters.set("depth", -1);
    parameters.set("penaltyDiscount", 4);
    parameters.set("useMaxPOrientationHeuristic", false);
    parameters.set("maxPOrientationMaxPathLength", 3);
    parameters.set("verbose", false);
    parameters.set("scaleFreeAlpha", 0.00001);
    parameters.set("scaleFreeBeta", 0.4);
    parameters.set("scaleFreeDeltaIn", .1);
    parameters.set("scaleFreeDeltaOut", 3);
    parameters.set("symmetricFirstStep", false);
    parameters.set("faithfulnessAssumed", true);
    parameters.set("maxDegree", 100);
    // parameters.set("logScale", true);
    Statistics statistics = new Statistics();
    statistics.add(new ParameterColumn("numMeasures"));
    statistics.add(new ParameterColumn("avgDegree"));
    statistics.add(new ParameterColumn("sampleSize"));
    statistics.add(new AdjacencyPrecision());
    statistics.add(new AdjacencyRecall());
    statistics.add(new ArrowheadPrecision());
    statistics.add(new ArrowheadRecall());
    statistics.add(new ElapsedTime());
    statistics.setWeight("AP", 0.25);
    statistics.setWeight("AR", 0.25);
    statistics.setWeight("AHP", 0.25);
    statistics.setWeight("AHR", 0.25);
    Algorithms algorithms = new Algorithms();
    Algorithm fges = new Fges(new SemBicScore());
    // algorithms.add(new FirstInflection(fges, "alpha", -7, -2, -.5));
    algorithms.add(new FirstInflection(fges, "penaltyDiscount", 0.7, 5, 1));
    Simulations simulations = new Simulations();
    simulations.add(new LinearFisherModel(new RandomForward()));
    Comparison comparison = new Comparison();
    comparison.setShowAlgorithmIndices(true);
    comparison.setShowSimulationIndices(true);
    comparison.setSortByUtility(false);
    comparison.setShowUtilities(false);
    comparison.setParallelized(true);
    comparison.setComparisonGraph(Comparison.ComparisonGraph.Pattern_of_the_true_DAG);
    comparison.compareFromSimulations("first.inflection", simulations, algorithms, statistics, parameters);
}

Example 2 with LinearFisherModel

use of edu.cmu.tetrad.algcomparison.simulation.LinearFisherModel in project tetrad by cmu-phil.

the class TestFges method main.

public static void main(String... args) {
    if (args.length > 0) {
        int numMeasures = Integer.parseInt(args[0]);
        int avgDegree = Integer.parseInt(args[1]);
        Parameters parameters = new Parameters();
        parameters.set("numMeasures", numMeasures);
        parameters.set("numLatents", 0);
        parameters.set("avgDegree", avgDegree);
        parameters.set("maxDegree", 20);
        parameters.set("maxIndegree", 20);
        parameters.set("maxOutdegree", 20);
        parameters.set("connected", false);
        parameters.set("coefLow", 0.2);
        parameters.set("coefHigh", 0.9);
        parameters.set("varLow", 1);
        parameters.set("varHigh", 3);
        parameters.set("verbose", false);
        parameters.set("coefSymmetric", true);
        parameters.set("numRuns", 1);
        parameters.set("percentDiscrete", 0);
        parameters.set("numCategories", 3);
        parameters.set("differentGraphs", true);
        parameters.set("sampleSize", 1000);
        parameters.set("intervalBetweenShocks", 10);
        parameters.set("intervalBetweenRecordings", 10);
        parameters.set("fisherEpsilon", 0.001);
        parameters.set("randomizeColumns", true);
        RandomGraph graph = new RandomForward();
        LinearFisherModel sim = new LinearFisherModel(graph);
        sim.createData(parameters);
        ScoreWrapper score = new FisherZScore();
        Algorithm alg = new edu.cmu.tetrad.algcomparison.algorithm.oracle.pattern.Fges(score);
        parameters.set("alpha", 1e-8);
        for (int i = 0; i < 5; i++) {
            Graph out1 = alg.search(sim.getDataModel(0), parameters);
            System.out.println(out1);
        }
    } else {
        new TestFges().test9();
    }
}

Example 3 with LinearFisherModel

use of edu.cmu.tetrad.algcomparison.simulation.LinearFisherModel in project tetrad by cmu-phil.

the class SimulationEditor method resetPanel.

private void resetPanel(Simulation simulation, String[] graphItems, String[] simulationItems, JTabbedPane tabbedPane) {
    RandomGraph randomGraph = (simulation.getSourceGraph() == null) ? new SingleGraph(new EdgeListGraph()) : new SingleGraph(simulation.getSourceGraph());
    if (!simulation.isFixedGraph()) {
        String graphItem = (String) graphsDropdown.getSelectedItem();
        simulation.getParams().set("graphsDropdownPreference", graphItem);
        if (graphItem.equals(graphItems[0])) {
            randomGraph = new RandomForward();
        } else if (graphItem.equals(graphItems[1])) {
            randomGraph = new ScaleFree();
        } else if (graphItem.equals(graphItems[2])) {
            randomGraph = new Cyclic();
        } else if (graphItem.equals(graphItems[3])) {
            randomGraph = new RandomSingleFactorMim();
        } else if (graphItem.equals(graphItems[4])) {
            randomGraph = new RandomTwoFactorMim();
        } else {
            throw new IllegalArgumentException("Unrecognized simulation type: " + graphItem);
        }
    }
    if (!simulation.isFixedSimulation()) {
        if (simulation.getSourceGraph() == null) {
            String simulationItem = (String) simulationsDropdown.getSelectedItem();
            simulation.getParams().set("simulationsDropdownPreference", simulationItem);
            simulation.setFixedGraph(false);
            if (randomGraph instanceof SingleGraph) {
                simulation.setFixedGraph(true);
            }
            if (simulationItem.equals(simulationItems[0])) {
                simulation.setSimulation(new BayesNetSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[1])) {
                simulation.setSimulation(new SemSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[2])) {
                simulation.setSimulation(new LinearFisherModel(randomGraph, simulation.getInputDataModelList()), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[3])) {
                simulation.setSimulation(new LeeHastieSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[4])) {
                simulation.setSimulation(new ConditionalGaussianSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[5])) {
                simulation.setSimulation(new TimeSeriesSemSimulation(randomGraph), simulation.getParams());
            } else {
                throw new IllegalArgumentException("Unrecognized simulation type: " + simulationItem);
            }
        } else {
            String simulationItem = (String) simulationsDropdown.getSelectedItem();
            simulation.getParams().set("simulationsDropdownPreference", simulationItem);
            simulation.setFixedGraph(false);
            if (randomGraph instanceof SingleGraph) {
                simulation.setFixedGraph(true);
            }
            if (simulationItem.equals(simulationItems[0])) {
                simulation.setSimulation(new BayesNetSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[1])) {
                simulation.setSimulation(new SemSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[2])) {
                simulation.setSimulation(new LinearFisherModel(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[3])) {
                simulation.setSimulation(new LeeHastieSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[4])) {
                simulation.setSimulation(new ConditionalGaussianSimulation(randomGraph), simulation.getParams());
            } else if (simulationItem.equals(simulationItems[5])) {
                simulation.setSimulation(new TimeSeriesSemSimulation(randomGraph), simulation.getParams());
            } else {
                throw new IllegalArgumentException("Unrecognized simulation type: " + simulationItem);
            }
        }
    }
    tabbedPane.setComponentAt(0, new PaddingPanel(getParameterPanel(simulation, simulation.getSimulation(), simulation.getParams())));
}

Example 4 with LinearFisherModel

use of edu.cmu.tetrad.algcomparison.simulation.LinearFisherModel in project tetrad by cmu-phil.

the class ExampleStARS method main.

public static void main(String... args) {
    Parameters parameters = new Parameters();
    // parameters.set("numMeasures", 100);
    // parameters.set("avgDegree", 2, 4);
    // parameters.set("sampleSize", 100, 500);
    // parameters.set("numRuns", 5);
    parameters.set("numMeasures", 200);
    parameters.set("avgDegree", 2, 4, 6);
    parameters.set("sampleSize", 100, 500);
    parameters.set("numRuns", 2);
    parameters.set("differentGraphs", true);
    parameters.set("numLatents", 0);
    parameters.set("maxDegree", 100);
    parameters.set("maxIndegree", 100);
    parameters.set("maxOutdegree", 100);
    parameters.set("connected", false);
    parameters.set("coefLow", 0.2);
    parameters.set("coefHigh", 0.9);
    parameters.set("varLow", 1);
    parameters.set("varHigh", 3);
    parameters.set("verbose", false);
    parameters.set("coefSymmetric", true);
    parameters.set("percentDiscrete", 0);
    parameters.set("numCategories", 3);
    parameters.set("differentGraphs", true);
    parameters.set("intervalBetweenShocks", 10);
    parameters.set("intervalBetweenRecordings", 10);
    parameters.set("fisherEpsilon", 0.001);
    parameters.set("randomizeColumns", true);
    parameters.set("alpha", 0.01);
    parameters.set("depth", -1);
    parameters.set("penaltyDiscount", 2);
    parameters.set("useMaxPOrientationHeuristic", false);
    parameters.set("maxPOrientationMaxPathLength", 3);
    parameters.set("verbose", false);
    parameters.set("scaleFreeAlpha", 0.00001);
    parameters.set("scaleFreeBeta", 0.4);
    parameters.set("scaleFreeDeltaIn", .1);
    parameters.set("scaleFreeDeltaOut", 3);
    parameters.set("symmetricFirstStep", false);
    parameters.set("faithfulnessAssumed", true);
    parameters.set("maxDegree", 100);
    parameters.set("StARS.tolerance", .5);
    parameters.set("StARS.cutoff", .05);
    parameters.set("numSubsamples", 7);
    parameters.set("percentSubsampleSize", .5);
    parameters.set("percentStability", .5);
    Statistics statistics = new Statistics();
    statistics.add(new ParameterColumn("numMeasures"));
    statistics.add(new ParameterColumn("avgDegree"));
    statistics.add(new ParameterColumn("sampleSize"));
    statistics.add(new AdjacencyPrecision());
    statistics.add(new AdjacencyRecall());
    statistics.add(new ArrowheadPrecision());
    statistics.add(new ArrowheadRecall());
    statistics.add(new ElapsedTime());
    statistics.setWeight("AP", 0.25);
    statistics.setWeight("AR", 0.25);
    statistics.setWeight("AHP", 0.25);
    statistics.setWeight("AHR", 0.25);
    Algorithms algorithms = new Algorithms();
    parameters.set("logScale", false);
    algorithms.add(new StabilitySelection(new Fges(new SemBicScore())));
    algorithms.add(new StARS(new Fges(new SemBicScore()), "penaltyDiscount", 1, 5));
    algorithms.add(new FirstInflection(new Fges(new SemBicScore()), "penaltyDiscount", 1, 5, .1));
    // parameters.set("penaltyDiscount", 5, 11, 15);
    // algorithms.add(new Fges(new SemBicScore()));
    // parameters.set("logScale", true);
    // Algorithm fges = new Fges(new FisherZScore());
    // algorithms.add(new StARS(fges, "alpha", -10, -2, -8));
    // algorithms.add(new FirstInflection(fges, "alpha", -10, -2, -3));
    Simulations simulations = new Simulations();
    simulations.add(new LinearFisherModel(new RandomForward()));
    Comparison comparison = new Comparison();
    comparison.setShowAlgorithmIndices(true);
    comparison.setShowSimulationIndices(true);
    comparison.setSortByUtility(false);
    comparison.setShowUtilities(false);
    comparison.setParallelized(true);
    comparison.setComparisonGraph(Comparison.ComparisonGraph.Pattern_of_the_true_DAG);
    comparison.compareFromSimulations("first.inflection", simulations, algorithms, statistics, parameters);
}

Example 5 with LinearFisherModel

use of edu.cmu.tetrad.algcomparison.simulation.LinearFisherModel in project tetrad by cmu-phil.

the class ExampleCompareSimulation method main.

public static void main(String... args) {
    if (args.length > 0) {
        int numMeasures = Integer.parseInt(args[0]);
        int avgDegree = Integer.parseInt(args[1]);
        Parameters parameters = new Parameters();
        // parameters.set("minCategories", 3);
        // parameters.set("maxCategories", 3);
        parameters.set("numRuns", 2);
        parameters.set("differentGraphs", true);
        parameters.set("sampleSize", 1000);
        parameters.set("numMeasures", numMeasures);
        parameters.set("numLatents", 0);
        parameters.set("avgDegree", avgDegree);
        parameters.set("maxDegree", 100);
        parameters.set("maxIndegree", 100);
        parameters.set("maxOutdegree", 100);
        parameters.set("connected", false);
        parameters.set("coefLow", 0.2);
        parameters.set("coefHigh", 0.9);
        parameters.set("varLow", 1);
        parameters.set("varHigh", 3);
        parameters.set("verbose", false);
        parameters.set("coefSymmetric", true);
        parameters.set("numRuns", 1);
        parameters.set("percentDiscrete", 0);
        parameters.set("numCategories", 3);
        parameters.set("differentGraphs", true);
        parameters.set("sampleSize", 1000);
        parameters.set("intervalBetweenShocks", 10);
        parameters.set("intervalBetweenRecordings", 10);
        parameters.set("fisherEpsilon", 0.001);
        parameters.set("randomizeColumns", true);
        parameters.set("alpha", 1e-8);
        parameters.set("depth", -1);
        parameters.set("penaltyDiscount", 4);
        parameters.set("useMaxPOrientationHeuristic", false);
        parameters.set("maxPOrientationMaxPathLength", 3);
        parameters.set("verbose", false);
        parameters.set("scaleFreeAlpha", 0.00001);
        parameters.set("scaleFreeBeta", 0.4);
        parameters.set("scaleFreeDeltaIn", .1);
        parameters.set("scaleFreeDeltaOut", 3);
        parameters.set("symmetricFirstStep", false);
        parameters.set("faithfulnessAssumed", true);
        parameters.set("maxDegree", 100);
        Statistics statistics = new Statistics();
        statistics.add(new ParameterColumn("numMeasures"));
        statistics.add(new ParameterColumn("avgDegree"));
        statistics.add(new AdjacencyPrecision());
        statistics.add(new AdjacencyRecall());
        statistics.add(new ArrowheadPrecision());
        statistics.add(new ArrowheadRecall());
        // statistics.add(new NumBidirectedEdges());
        // statistics.add(new MathewsCorrAdj());
        // statistics.add(new MathewsCorrArrow());
        // statistics.add(new F1Adj());
        // statistics.add(new F1Arrow());
        // statistics.add(new SHD());
        statistics.add(new ElapsedTime());
        statistics.setWeight("AP", 0.25);
        statistics.setWeight("AR", 0.25);
        statistics.setWeight("AHP", 0.25);
        statistics.setWeight("AHR", 0.25);
        Algorithms algorithms = new Algorithms();
        // algorithms.add(new Pc(new FisherZ()));
        // algorithms.add(new Pc(new SemBicTest()));
        // algorithms.add(new Cpc(new FisherZ()));
        // algorithms.add(new PcStable(new FisherZ()));
        // algorithms.add(new CpcStable(new FisherZ()));
        // algorithms.add(new PcStableMax(new FisherZ(), false));
        // algorithms.add(new PcStableMax(new SemBicTest(), false));
        algorithms.add(new Fges(new FisherZScore(), false));
        // algorithms.add(new Fges(new SemBicScore(), false));
        Simulations simulations = new Simulations();
        simulations.add(new LinearFisherModel(new RandomForward()));
        Comparison comparison = new Comparison();
        comparison.setShowAlgorithmIndices(true);
        comparison.setShowSimulationIndices(true);
        comparison.setSortByUtility(false);
        comparison.setShowUtilities(true);
        comparison.compareFromSimulations("comparisonJoe", simulations, algorithms, statistics, parameters);
    }
}