tasksimulators.newsimu.AGL
Class AgentAGL

java.lang.Object
  extended byclarion.system.TaskAgent
      extended bytasksimulators.newsimu.AGL.AgentAGL

public class AgentAGL
extends TaskAgent


Field Summary
 
Fields inherited from class clarion.system.TaskAgent
activeAmnIdx, clarion, curNacsInput, curSensoryInput, desiredOutput, global, learningTestOn, nacsResults, reinforcements, taskClarion
 
Constructor Summary
AgentAGL()
           
AgentAGL(Global g, short[][][] trainSet, short[][][] testSet)
           
 
Method Summary
 short[] decideFrAction(int netIdx, int setIdx, short[] input)
           
 short[] decideIrlAction(int netIdx, int setIdx, int ruleNo, short[] input)
           
 int determineAmnNet(int netIdx, short[][][] action)
          Determines which AMN will be used given current action.
 void getAccuracy(double[] arr)
          Get the data of Accuracy and populate the array.
 void getAchievement(double[] arr)
          Get the data of achievement and populate the array.
 void getAttempts(double[] arr)
          Get the data of Attempts and populate the array.
 java.lang.String getCurSensoryInputLabel()
           
 int getCurSensoryInputType()
           
 void getEfficiency(double[] arr)
          Get the data of Efficiency and populate the array.
 void getOutput(java.lang.Object arr)
          Returns the output of current step.
 void getSpeed(double[] arr)
          Get the data of Speed and populate the array.
 boolean isControlNacsAction(int netIdx, short[] chosenAction)
           
 void performAction(int netIdx, short[][][] action)
          perform the selected action from a particular ACS network.
 void reinit()
          reinitialization routine.
 void run()
          Implements the overall process of a subject during the AGL task experiment.
 void setGroupType(char[] groupType)
           
 void setRetrievalResults(int netIdx, GKSChunk[] results)
          send the retrieval results from NACS to ACS components.
 
Methods inherited from class clarion.system.TaskAgent
getActiveAmnIdx, getCurDesiredOutput, getCurNacsInput, getCurSensoryInput, getFixRules, getIrlRules, getRerRules, isInLearningTest, saveOutput, setLearningTest
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

AgentAGL

public AgentAGL()

AgentAGL

public AgentAGL(Global g,
                short[][][] trainSet,
                short[][][] testSet)
Method Detail

reinit

public void reinit()
Description copied from class: TaskAgent
reinitialization routine.

Overrides:
reinit in class TaskAgent

isControlNacsAction

public boolean isControlNacsAction(int netIdx,
                                   short[] chosenAction)

getAchievement

public void getAchievement(double[] arr)
Get the data of achievement and populate the array.

Parameters:
arr - the array to store the data.

getAccuracy

public void getAccuracy(double[] arr)
Get the data of Accuracy and populate the array.

Parameters:
arr - the array to store the data.

getEfficiency

public void getEfficiency(double[] arr)
Get the data of Efficiency and populate the array.

Parameters:
arr - the array to store the data.

getSpeed

public void getSpeed(double[] arr)
Get the data of Speed and populate the array.

Parameters:
arr - the array to store the data.

getAttempts

public void getAttempts(double[] arr)
Get the data of Attempts and populate the array.

Parameters:
arr - the array to store the data.

getOutput

public void getOutput(java.lang.Object arr)
Description copied from class: TaskAgent
Returns the output of current step.

Specified by:
getOutput in class TaskAgent
Parameters:
arr - the array to be filled in with current output.

getCurSensoryInputType

public int getCurSensoryInputType()

getCurSensoryInputLabel

public java.lang.String getCurSensoryInputLabel()

setRetrievalResults

public void setRetrievalResults(int netIdx,
                                GKSChunk[] results)
Description copied from class: TaskAgent
send the retrieval results from NACS to ACS components.

Specified by:
setRetrievalResults in class TaskAgent
Parameters:
netIdx - the ACS network index.
results - retrieval results from NACS

setGroupType

public void setGroupType(char[] groupType)

decideIrlAction

public short[] decideIrlAction(int netIdx,
                               int setIdx,
                               int ruleNo,
                               short[] input)

decideFrAction

public short[] decideFrAction(int netIdx,
                              int setIdx,
                              short[] input)

determineAmnNet

public int determineAmnNet(int netIdx,
                           short[][][] action)
Description copied from class: TaskAgent
Determines which AMN will be used given current action.

Specified by:
determineAmnNet in class TaskAgent
Parameters:
netIdx - the ACS network index.
action - the selected actions by ACS networks.
Returns:
the AMN network index.

performAction

public void performAction(int netIdx,
                          short[][][] action)
perform the selected action from a particular ACS network.

Overrides:
performAction in class TaskAgent
Parameters:
netIdx - the ACS network index
action - the selected action from the specified ACS network.

run

public void run()
Implements the overall process of a subject during the AGL task experiment. It is assumed to be called by the TaskAGL class to implement what a subject have done during the experiment.

Specified by:
run in class TaskAgent