Class TrainedDagSimulatorGNM2.NodeReport

java.lang.Object
edu.cmu.tetrad.sem.TrainedDagSimulatorGNM2.NodeReport
Enclosing class:
TrainedDagSimulatorGNM2
public static final class TrainedDagSimulatorGNM2.NodeReport extends Object
Represents a report for a single node in a trained directed acyclic graph (DAG) structure. The report contains metrics specific to the node, which can vary depending on whether the node represents a discrete or continuous variable. This is an immutable class.

  • Field Summary

    Fields
    Modifier and Type
    Field
    Description
    final boolean
    discreteChild
    Whether the node represents a discrete variable.
    final double
    mseTrain
    Mean Squared Error (MSE) for the training dataset associated with the node.
    final String
    node
    Node identifier.
    final int
    numLevels
    Number of discrete levels for the node.
    final List<String>
    parents
    List of parent nodes for the current node.
    final double
    residualMean
    Mean and standard deviation of residuals for the training dataset associated with the node.
    final double
    residualSd
    Standard deviation of residuals for the training dataset associated with the node.
    final int
    trainingRowsUsed
    Number of training rows used for the node.
    final double
    xentTrain
    Cross-entropy loss for the training dataset associated with the node.

  • Method Summary

    Methods inherited from class Object

    equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

  • Field Details

    • node

      public final String node
      Node identifier.

    • discreteChild

      public final boolean discreteChild
      Whether the node represents a discrete variable.

    • parents

      public final List<String> parents
      List of parent nodes for the current node.

    • trainingRowsUsed

      public final int trainingRowsUsed
      Number of training rows used for the node.

    • mseTrain

      public final double mseTrain
      Mean Squared Error (MSE) for the training dataset associated with the node. This metric is relevant only for nodes representing continuous variables. If the node is a discrete variable, this value will be NaN.

    • residualMean

      public final double residualMean
      Mean and standard deviation of residuals for the training dataset associated with the node. This metric is relevant only for nodes representing continuous variables. If the node is a discrete variable, these values will be NaN.

    • residualSd

      public final double residualSd
      Standard deviation of residuals for the training dataset associated with the node. This metric is relevant only for nodes representing continuous variables. If the node is a discrete variable, this value will be NaN.

    • xentTrain

      public final double xentTrain
      Cross-entropy loss for the training dataset associated with the node. This metric is relevant only for nodes representing discrete variables. If the node is a continuous variable, this value will be NaN.

    • numLevels

      public final int numLevels
      Number of discrete levels for the node. This metric is relevant only for nodes representing discrete variables. If the node is a continuous variable, this value will be 0.