Package jline.examples.parity

Class StateProbTwins

java.lang.Object
jline.examples.parity.StateProbTwins
public final class StateProbTwins extends Object
python/examples/advanced/stateProbabilities/: the four state-probability queries, getProbAggr, getProb, getProbSysAggr and getProbSys.

WHAT A station.setState(n) MEANS HERE. The reference scripts set a state on the model and then ask a solver for its probability, and the JAR carries that state on the StatefulNode itself, so the twin does exactly what the reference does: write the per-class counts onto the stations the reference writes them onto, then query. A station the reference leaves at -1 is IGNORED by the query it feeds (getProbAggr reads one station's row, getProbSysAggr reads them all) and is left at the model's own default initialization.

THE GOLDEN IS THE BARE SCALAR. Each of these examples is goldened on one number, the CTMC probability, which the reference prints on a line of its own after a Pmarg_ctmc = label; parse_optimization_scalar keeps the FIRST bare float in (0, 100). Everything else printed here is labelled, so the twin can report the whole comparison without moving that cell.

  • Method Summary

    Modifier and Type
    Method
    Description
    static void
    allprobs(Network model, int samples)
    statepr_allprobs_fcfs.py / statepr_allprobs_ps.py: the four queries on the two-class class-switching model, under every solver that carries them.
    static void
    stateprAggr()
    statepr_aggr.py: the marginal aggregate probability of one station, by the exact chain and by the normalizing constants.
    static void
    stateprAggrLarge()
    statepr_aggr_large.py: the same query on a four-class model whose class switches all happen at Queue2.
    static void
    sysAggr(Network model, double[] last)
    statepr_sys_aggr.py: the joint aggregate probability that every job sits at the last station.
    static void
    sysAggrLarge()
    statepr_sys_aggr_large.py: the same joint query on three queues and four unit-population classes.

    Methods inherited from class java.lang.Object

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

  • Method Details

    • stateprAggr

      public static void stateprAggr()
      statepr_aggr.py: the marginal aggregate probability of one station, by the exact chain and by the normalizing constants.

    • stateprAggrLarge

      public static void stateprAggrLarge()
      statepr_aggr_large.py: the same query on a four-class model whose class switches all happen at Queue2.

    • allprobs

      public static void allprobs(Network model, int samples) throws Exception
      statepr_allprobs_fcfs.py / statepr_allprobs_ps.py: the four queries on the two-class class-switching model, under every solver that carries them.
      Throws:
      Exception

    • sysAggr

      public static void sysAggr(Network model, double[] last)
      statepr_sys_aggr.py: the joint aggregate probability that every job sits at the last station.

    • sysAggrLarge

      public static void sysAggrLarge()
      statepr_sys_aggr_large.py: the same joint query on three queues and four unit-population classes. The MATLAB reference reports about 0.000348.