Examples with DefaultDataset com.sri.ai.praise.learning.parameterlearning.representation.dataset.DefaultDataset used on opensource projects

Example 6 with DefaultDataset

use of com.sri.ai.praise.learning.parameterlearning.representation.dataset.DefaultDataset in project aic-praise by aic-sri-international.

the class TableBayesianModelTest method testSickSunColdModelWithDifferentDatapoints.

@Test
public void testSickSunColdModelWithDifferentDatapoints() {
    // Dataset
    List<TableVariable> variables = list(sickVariable, sunVariable, coldVariable);
    List<Integer> variableValues1 = list(1, 0, 1);
    DefaultDatapoint datapoint1 = new DefaultDatapoint(variables, variableValues1);
    List<DefaultDatapoint> datapoints = list();
    int numberOfDatapoints1 = 4;
    for (int i = 1; i <= numberOfDatapoints1; i++) {
        datapoints.add(datapoint1);
    }
    List<Integer> variableValues2 = list(0, 0, 0);
    DefaultDatapoint datapoint2 = new DefaultDatapoint(variables, variableValues2);
    int numberOfDatapoints2 = 2;
    for (int i = 1; i <= numberOfDatapoints2; i++) {
        datapoints.add(datapoint2);
    }
    List<Integer> variableValues3 = list(0, 0, 1);
    DefaultDatapoint datapoint3 = new DefaultDatapoint(variables, variableValues3);
    int numberOfDatapoints3 = 1;
    for (int i = 1; i <= numberOfDatapoints3; i++) {
        datapoints.add(datapoint3);
    }
    DefaultDataset dataset = new DefaultDataset(datapoints);
    // Learning
    sickSunColdModel = (TableBayesianModel) sickSunColdModel.learnModelParametersFromCompleteData(dataset);
    List<? extends TableBayesianNode> learnedNodes = sickSunColdModel.getNodes();
    // Testing
    // For the sickNode first:
    // Expected parameters (2 datapoints1): {(0, [0, 0])=0.5, (1, [0, 0])=0.5, (1, [1, 0])=0.5, (1, [1, 1])=0.5, (0, [1, 1])=0.5, (0, [1, 0])=0.5, (0, [0, 1])=0.25, (1, [0, 1])=0.75}
    TableBayesianNode learnedSickNode = learnedNodes.get(0);
    LinkedHashMap<TableVariable, Integer> variablesAndTheirValues = map();
    variablesAndTheirValues.put(sickVariable, 0);
    variablesAndTheirValues.put(sunVariable, 0);
    variablesAndTheirValues.put(coldVariable, 0);
    // Parameter for (0, [0, 0]):
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints2) / (2 + numberOfDatapoints2)), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [0, 0]):
    variablesAndTheirValues.put(sickVariable, 1);
    Assert.assertEquals(Double.valueOf(1.0 / (2 + numberOfDatapoints2)), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [1, 0]):
    variablesAndTheirValues.put(sunVariable, 1);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [1, 1]):
    variablesAndTheirValues.put(coldVariable, 1);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (0, [1, 1]):
    variablesAndTheirValues.put(sickVariable, 0);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (0, [1, 0]):
    variablesAndTheirValues.put(coldVariable, 0);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (0, [0, 1]):
    variablesAndTheirValues.put(sunVariable, 0);
    variablesAndTheirValues.put(coldVariable, 1);
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints3) / (2 + numberOfDatapoints1 + numberOfDatapoints3)), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [0, 1]):
    variablesAndTheirValues.put(sickVariable, 1);
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints1) / (2 + numberOfDatapoints1 + numberOfDatapoints3)), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // For the sunNode:
    // Expected parameters (2 datapoints): {(0, [])=0.75, (1, [])=0.25}
    TableBayesianNode learnedSunNode = learnedNodes.get(1);
    variablesAndTheirValues = map();
    variablesAndTheirValues.put(sunVariable, 0);
    // Parameter for (0, []):
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints1 + numberOfDatapoints2 + numberOfDatapoints3) / (2 + numberOfDatapoints1 + numberOfDatapoints2 + numberOfDatapoints3)), learnedSunNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, []):
    variablesAndTheirValues.put(sunVariable, 1);
    Assert.assertEquals(Double.valueOf(1.0 / (2 + numberOfDatapoints1 + numberOfDatapoints2 + numberOfDatapoints3)), learnedSunNode.getEntryFor(variablesAndTheirValues));
    // For the coldNode:
    // Expected parameters (2 datapoints): {(0, [])=0.25, (1, [])=0.75}
    TableBayesianNode learnedColdNode = learnedNodes.get(2);
    variablesAndTheirValues = map();
    variablesAndTheirValues.put(coldVariable, 0);
    // Parameter for (0, []):
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints2) / (2 + numberOfDatapoints1 + numberOfDatapoints2 + numberOfDatapoints3)), learnedColdNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, []):
    variablesAndTheirValues.put(coldVariable, 1);
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints1 + numberOfDatapoints3) / (2 + numberOfDatapoints1 + numberOfDatapoints2 + numberOfDatapoints3)), learnedColdNode.getEntryFor(variablesAndTheirValues));
}

Example 7 with DefaultDataset

use of com.sri.ai.praise.learning.parameterlearning.representation.dataset.DefaultDataset in project aic-praise by aic-sri-international.

the class TableBayesianModelTest method testSickSunColdModel.

@Test
public void testSickSunColdModel() {
    // Dataset
    List<TableVariable> variables = list(sickVariable, sunVariable, coldVariable);
    List<Integer> variableValues = list(1, 0, 1);
    DefaultDatapoint datapoint = new DefaultDatapoint(variables, variableValues);
    List<DefaultDatapoint> datapoints = list();
    int numberOfDatapoints = 2;
    for (int i = 1; i <= numberOfDatapoints; i++) {
        datapoints.add(datapoint);
    }
    DefaultDataset dataset = new DefaultDataset(datapoints);
    // Learning
    sickSunColdModel = (TableBayesianModel) sickSunColdModel.learnModelParametersFromCompleteData(dataset);
    List<? extends TableBayesianNode> learnedNodes = sickSunColdModel.getNodes();
    // Testing
    // For the sickNode first:
    // Expected parameters (2 datapoints): {(0, [0, 0])=0.5, (1, [0, 0])=0.5, (1, [1, 0])=0.5, (1, [1, 1])=0.5, (0, [1, 1])=0.5, (0, [1, 0])=0.5, (0, [0, 1])=0.25, (1, [0, 1])=0.75}
    TableBayesianNode learnedSickNode = learnedNodes.get(0);
    LinkedHashMap<TableVariable, Integer> variablesAndTheirValues = map();
    variablesAndTheirValues.put(sickVariable, 0);
    variablesAndTheirValues.put(sunVariable, 0);
    variablesAndTheirValues.put(coldVariable, 0);
    // Parameter for (0, [0, 0]):
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [0, 0]):
    variablesAndTheirValues.put(sickVariable, 1);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [1, 0]):
    variablesAndTheirValues.put(sunVariable, 1);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [1, 1]):
    variablesAndTheirValues.put(coldVariable, 1);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (0, [1, 1]):
    variablesAndTheirValues.put(sickVariable, 0);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (0, [1, 0]):
    variablesAndTheirValues.put(coldVariable, 0);
    Assert.assertEquals(Double.valueOf(0.5), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (0, [0, 1]):
    variablesAndTheirValues.put(sunVariable, 0);
    variablesAndTheirValues.put(coldVariable, 1);
    Assert.assertEquals(Double.valueOf(1.0 / (2 + numberOfDatapoints)), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, [0, 1]):
    variablesAndTheirValues.put(sickVariable, 1);
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints) / (2 + numberOfDatapoints)), learnedSickNode.getEntryFor(variablesAndTheirValues));
    // For the sunNode:
    // Expected parameters (2 datapoints): {(0, [])=0.75, (1, [])=0.25}
    TableBayesianNode learnedSunNode = learnedNodes.get(1);
    variablesAndTheirValues = map();
    variablesAndTheirValues.put(sunVariable, 0);
    // Parameter for (0, []):
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints) / (2 + numberOfDatapoints)), learnedSunNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, []):
    variablesAndTheirValues.put(sunVariable, 1);
    Assert.assertEquals(Double.valueOf(1.0 / (2 + numberOfDatapoints)), learnedSunNode.getEntryFor(variablesAndTheirValues));
    // For the coldNode:
    // Expected parameters (2 datapoints): {(0, [])=0.25, (1, [])=0.75}
    TableBayesianNode learnedColdNode = learnedNodes.get(2);
    variablesAndTheirValues = map();
    variablesAndTheirValues.put(coldVariable, 0);
    // Parameter for (0, []):
    Assert.assertEquals(Double.valueOf(1.0 / (2 + numberOfDatapoints)), learnedColdNode.getEntryFor(variablesAndTheirValues));
    // Parameter for (1, []):
    variablesAndTheirValues.put(coldVariable, 1);
    Assert.assertEquals(Double.valueOf((1.0 + numberOfDatapoints) / (2 + numberOfDatapoints)), learnedColdNode.getEntryFor(variablesAndTheirValues));
}

Example 8 with DefaultDataset

use of com.sri.ai.praise.learning.parameterlearning.representation.dataset.DefaultDataset in project aic-praise by aic-sri-international.

the class TableBayesianModelTest method printSickSunColdModelTest.

public static void printSickSunColdModelTest() {
    // Dataset
    List<TableVariable> variables = list(sickVariable, sunVariable, coldVariable);
    List<Integer> variableValues = list(1, 0, 1);
    DefaultDatapoint datapoint = new DefaultDatapoint(variables, variableValues);
    List<DefaultDatapoint> datapoints = list();
    int numberOfDatapoints = 2;
    for (int i = 1; i <= numberOfDatapoints; i++) {
        datapoints.add(datapoint);
    }
    DefaultDataset dataset = new DefaultDataset(datapoints);
    // Learning
    long startTime = System.currentTimeMillis();
    sickSunColdModel = (TableBayesianModel) sickSunColdModel.learnModelParametersFromCompleteData(dataset);
    long stopTime = System.currentTimeMillis();
    long elapsedTime = stopTime - startTime;
    System.out.println("Elapsed time for learning with " + numberOfDatapoints + " datapoints: " + elapsedTime + " miliseconds \n");
    List<? extends TableBayesianNode> learnedNodes = sickSunColdModel.getNodes();
    // Testing
    String expectedParametersForSick = "{(0, [0, 0])=0.5, (1, [0, 0])=0.5, (0, [0, 1])=0.25, (1, [0, 1])=0.75, (0, [1, 0])=0.5, (1, [1, 0])=0.5, (0, [1, 1])=0.5, (1, [1, 1])=0.5}";
    System.out.println("Expected parameters for sick (with 2 datapoints):\n" + expectedParametersForSick + "\n");
    TableBayesianNode learnedSickNode = learnedNodes.get(0);
    LinkedHashMap<TableVariable, Integer> variablesAndTheirValues = new LinkedHashMap<TableVariable, Integer>();
    variablesAndTheirValues.put(sickVariable, 1);
    variablesAndTheirValues.put(sunVariable, 0);
    variablesAndTheirValues.put(coldVariable, 1);
    System.out.println("Actual entries for sick:");
    System.out.println("entryFor(" + variablesAndTheirValues.get(sickVariable) + ", [" + variablesAndTheirValues.get(sunVariable) + ", " + variablesAndTheirValues.get(coldVariable) + "]) = " + learnedSickNode.getEntryFor(variablesAndTheirValues));
    variablesAndTheirValues.put(sickVariable, 0);
    System.out.println("entryFor(" + variablesAndTheirValues.get(sickVariable) + ", [" + variablesAndTheirValues.get(sunVariable) + ", " + variablesAndTheirValues.get(coldVariable) + "]) = " + learnedSickNode.getEntryFor(variablesAndTheirValues));
    variablesAndTheirValues.put(coldVariable, 0);
    System.out.println("entryFor(" + variablesAndTheirValues.get(sickVariable) + ", [" + variablesAndTheirValues.get(sunVariable) + ", " + variablesAndTheirValues.get(coldVariable) + "]) = " + learnedSickNode.getEntryFor(variablesAndTheirValues));
}

Example 9 with DefaultDataset

use of com.sri.ai.praise.learning.parameterlearning.representation.dataset.DefaultDataset in project aic-praise by aic-sri-international.

the class ExpressionBayesianModelTest method testEarthquakeBurglaryAlarmModel.

/**
 * For the alarmNode:
 * We have four families (one for every (Earthquake, Burglary) pair value), and naturally each one with two parameters (one for Alarm = 1 and the other for Alarm = 0)
 *
 * Final families:
 * F1: [Condition: Earthquake = 1 and Burglary = 1, Parameters: [Param1, OneMinusParam1]]
 * F2: [Condition: Earthquake = 1 and Burglary = 0, Parameters: [Param2, OneMinusParam2]]
 * F3: [Condition: Earthquake = 0 and Burglary = 1, Parameters: [Param3, OneMinusParam3]]
 * F4: [Condition: Earthquake = 0 and Burglary = 0, Parameters: [Param4, OneMinusParam4]]
 *
 * For the burglaryNode:
 * Similar and much simpler, we have only one family here, with two parameters (one for Burglary = 1 and the other for Burglary = 0)
 * F1burglary = [Condition: true, Parameters: [Param5, OneMinusParam5]]
 *
 * For the earthquakeNode:
 * Since we have no parameters to learn we have no families here, this expressionFactor is treated as a constant prior probability defined by the user (here, P(Earthquake) is set to 1% in the generateEarthquakeBurglaryAlarmModel method)
 */
@Test
public void testEarthquakeBurglaryAlarmModel() {
    ExpressionBayesianModel model = generateEarthquakeBurglaryAlarmModel();
    // Dataset - Order of variables for the datapoints: (Alarm, Earthquake, Burglary)
    // (1, 0, 1)
    int numberOfDatapoints1 = 1;
    // (1, 1, 1)
    int numberOfDatapoints2 = 2;
    DefaultDataset dataset = generateDatasetForEarthquakeBurglaryAlarmModel(numberOfDatapoints1, numberOfDatapoints2);
    // Expected values for alarmNode
    Expression expectedParam3inF3 = parse("(1 + " + numberOfDatapoints1 + ") / (2 + " + numberOfDatapoints1 + ")");
    Expression expectedOneMinusParam3inF3 = parse("1 / (2 + " + numberOfDatapoints1 + ")");
    Expression expectedParam1inF1 = parse("(1 + " + numberOfDatapoints2 + ") / (2 + " + numberOfDatapoints2 + ")");
    Expression expectedOneMinusParam1inF1 = parse("1 / (2 + " + numberOfDatapoints2 + ")");
    Expression expectedAlarmNode = parse("if Earthquake = 0 and Burglary = 1 then if Alarm = 1 then " + expectedParam3inF3 + " else " + expectedOneMinusParam3inF3 + " else if Earthquake = 1 and Burglary = 1 then if Alarm = 1 then " + expectedParam1inF1 + " else " + expectedOneMinusParam1inF1 + " else 0.5");
    // Expected values for burglaryNode
    Expression expectedParam5inF1burglary = parse("(1 + " + (numberOfDatapoints1 + numberOfDatapoints2) + ") / (2 + " + (numberOfDatapoints1 + numberOfDatapoints2) + ")");
    Expression expectedOneMinusParam5inF1burglary = parse("1 / (2 + " + (numberOfDatapoints1 + numberOfDatapoints2) + ")");
    Expression expectedBurglaryNode = parse("if Burglary = 1 then " + expectedParam5inF1burglary + " else " + expectedOneMinusParam5inF1burglary);
    // Expected value for earthquakeNode (fixed prior probability)
    Expression expectedEarthquakeNode = parse("if Earthquake = 1 then 0.01 else 0.99");
    // Learning
    model = (ExpressionBayesianModel) model.learnModelParametersFromCompleteData(dataset);
    ExpressionBayesianNode learnedAlarmNode = model.getNodes().get(0);
    ExpressionBayesianNode learnedEarthquakeNode = model.getNodes().get(1);
    ExpressionBayesianNode learnedBurglaryNode = model.getNodes().get(2);
    // Verifications
    Expression alarmNodeVerification = Equality.make(expectedAlarmNode, learnedAlarmNode);
    Expression earthquakeNodeVerification = Equality.make(expectedEarthquakeNode, learnedEarthquakeNode);
    Expression burglaryNodeVerification = Equality.make(expectedBurglaryNode, learnedBurglaryNode);
    alarmNodeVerification = contextForEarthquakeBurglaryAlarmModel.evaluate(alarmNodeVerification);
    earthquakeNodeVerification = contextForEarthquakeBurglaryAlarmModel.evaluate(earthquakeNodeVerification);
    burglaryNodeVerification = contextForEarthquakeBurglaryAlarmModel.evaluate(burglaryNodeVerification);
    // Assertions
    assertEquals(Expressions.TRUE, alarmNodeVerification);
    assertEquals(Expressions.TRUE, earthquakeNodeVerification);
    assertEquals(Expressions.TRUE, burglaryNodeVerification);
}

Example 10 with DefaultDataset

use of com.sri.ai.praise.learning.parameterlearning.representation.dataset.DefaultDataset in project aic-praise by aic-sri-international.

the class ExpressionBayesianModelTest method testChildParentModel1.

/**
 * Only one family having the two parameters
 * expressionForChildNode: if Child < 5 then Param1 else Param2
 *
 * Final families:
 * F1: [Condition = true, Parameters: [Param1, Param2]]
 */
@Test
public void testChildParentModel1() {
    ExpressionBayesianModel model = generateChildParentModel(parse("if Child < 5 then Param1 else Param2"));
    // (1, 2)
    int numberOfDatapoints1 = 1;
    // (5, 1)
    int numberOfDatapoints2 = 2;
    DefaultDataset dataset = generateDatasetForChildParentModel(numberOfDatapoints1, numberOfDatapoints2, 0, 0);
    model = (ExpressionBayesianModel) model.learnModelParametersFromCompleteData(dataset);
    ExpressionBayesianNode learnedChild = model.getNodes().get(0);
    ExpressionBayesianNode learnedParent = model.getNodes().get(1);
    Expression expectedParam1inF1 = parse("( (4 + " + numberOfDatapoints1 + ")/(5 + " + (numberOfDatapoints1 + numberOfDatapoints2) + ") ) / 4");
    Expression expectedParam2inF1 = parse("(1 + " + numberOfDatapoints2 + ")/(5 + " + (numberOfDatapoints1 + numberOfDatapoints2) + ")");
    Expression expectedChildExpression = parse("if Child < 5 then " + expectedParam1inF1 + " else " + expectedParam2inF1);
    Expression expectedParentExpression = parse("0.2");
    Expression childVerification = Equality.make(expectedChildExpression, learnedChild);
    Expression parentVerification = Equality.make(expectedParentExpression, learnedParent);
    // println(childVerification); // uncomment this line if you want to see the main equality that is being tested
    childVerification = contextForChildParentModel.evaluate(childVerification);
    parentVerification = contextForChildParentModel.evaluate(parentVerification);
    assertEquals(Expressions.TRUE, childVerification);
    assertEquals(Expressions.TRUE, parentVerification);
}