Package edu.cmu.tetrad.sem

Class SemOptimizerRegression

java.lang.Object
edu.cmu.tetrad.sem.SemOptimizerRegression
All Implemented Interfaces:
SemOptimizer, TetradSerializable, Serializable
public class SemOptimizerRegression extends Object implements SemOptimizer
Optimizes a DAG SEM with no hidden variables using closed formula regressions. IT SHOULD NOT BE USED WITH SEMs THAT ARE NOT DAGS OR CONTAIN HIDDEN NODES. IT ALSO ASSUMES THAT ALL OBSERVED NODES APPEAR FIRST IN semIm.getSemPm().getDag().getNodes(), I.E., ERROR NODES ARE INSERTED ONLY AFTER MEASURED NODES IN THIS LIST.
Author:
Ricardo Silva
See Also:

  • Constructor Details

    • SemOptimizerRegression

      public SemOptimizerRegression()
      Blank constructor.

  • Method Details

    • serializableInstance

      public static SemOptimizerRegression serializableInstance()
      Generates a simple exemplar of this class to test serialization.

    • optimize

      public void optimize(SemIm semIm)
      Fit the freeParameters by doing local regressions.
      Specified by:
      optimize in interface SemOptimizer
      Parameters:
      semIm - The unoptimized Sem (will be optimized).

    • setNumRestarts

      public void setNumRestarts(int numRestarts)
      Specified by:
      setNumRestarts in interface SemOptimizer

    • getNumRestarts

      public int getNumRestarts()
      Specified by:
      getNumRestarts in interface SemOptimizer

    • toString

      public String toString()
      Overrides:
      toString in class Object