org.joone.io
Class StreamOutputSynapse

java.lang.Object
  extended by org.joone.engine.Synapse
      extended by org.joone.io.StreamOutputSynapse
All Implemented Interfaces:
java.io.Serializable, InputPatternListener, Learnable, LearnableSynapse, NeuralElement, OutputPatternListener, Inspectable, PlugInListener
Direct Known Subclasses:
FileOutputSynapse, ImageOutputSynapse, JDBCOutputSynapse, MemoryOutputSynapse, XLSOutputSynapse

public abstract class StreamOutputSynapse
extends Synapse
implements PlugInListener

See Also:
Serialized Form

Field Summary
protected  Fifo fifo
           
protected  OutputConverterPlugIn nextPlugIn
           
 
Fields inherited from class org.joone.engine.Synapse
array, b_pattern, bitems, bouts, count, enabled, fwdLock, inps, items, learnable, m_batch, m_pattern, myLearner, outs, revLock
 
Constructor Summary
StreamOutputSynapse()
           
 
Method Summary
 boolean addPlugIn(OutputConverterPlugIn aNewPlugIn)
          Adds a plug in to the stream output synapse for data preprocessing.
protected  void backward(double[] pattern)
          Funzione di TRAIN dell'elemento.
 java.util.TreeSet check()
          Base for check messages.
 void dataChanged(PlugInEvent data)
          This method is called by plug-ins whenever data is changed.
protected  void forward(double[] pattern)
          Recall function
 void fwdPut(Pattern pattern)
          The standard fwdPut method.
protected  Fifo getFifo()
          Getter for property fifo.
 OutputConverterPlugIn getPlugIn()
           
 char getSeparator()
          Returns the column separator creation date: (23/04/00 0.50.18)
 boolean isBuffered()
          Checks the buffered status of this synapse.
 void removeAllPlugIns()
          Removes (and disconnects) all (cascading) plug-ins.
 Pattern revGet()
          Returns the error pattern coming from the next layer during the training phase
protected  void setArrays(int rows, int cols)
          setArrays method comment.
 void setBuffered(boolean buf)
          Sets the buffered status of this synapse.
protected  void setDimensions(int rows, int cols)
          Dimensiona l'elemento
 boolean setPlugIn(OutputConverterPlugIn newPlugIn)
          Deprecated. use addPlugIn(OutputConverterPlugIn)
 void setSeparator(char newSeparator)
          Inserire qui la descrizione del metodo.
abstract  void write(Pattern pattern)
          Custom xxxOutputSynapses need to implement at least this method.
 
Methods inherited from class org.joone.engine.Synapse
addNoise, canCountSteps, fwdGet, fwdPattern_consume, fwdPattern_produce, getCount, getFwdLock, getIgnoreBefore, getInputDimension, getLearner, getLearningRate, getMomentum, getMonitor, getName, getOutputDimension, getRevLock, getSynapseNameSafely, getWeights, init, initLearner, InspectableTitle, Inspections, isEnabled, isInputFull, isLoopBack, isOutputFull, randomize, readResolve, reset, revPattern_consume, revPattern_produce, revPut, setEnabled, setIgnoreBefore, setInputDimension, setInputFull, setLearningRate, setLoopBack, setMomentum, setMonitor, setName, setOutputDimension, setOutputFull, setWeights, warnLogger, warnLogger
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

fifo

protected transient Fifo fifo

nextPlugIn

protected OutputConverterPlugIn nextPlugIn
Constructor Detail

StreamOutputSynapse

public StreamOutputSynapse()
Method Detail

backward

protected void backward(double[] pattern)
Description copied from class: Synapse
Funzione di TRAIN dell'elemento.

Specified by:
backward in class Synapse
Parameters:
pattern - double[] - pattern di input sul quale applicare la funzione di trasferimento

forward

protected void forward(double[] pattern)
Description copied from class: Synapse
Recall function

Specified by:
forward in class Synapse
Parameters:
pattern - double[] - input pattern. THIS PATTERN CANNOT BE CHANGED

fwdPut

public void fwdPut(Pattern pattern)
The standard fwdPut method. This method performs buffering of pattern data if required according to the buffered flag. If buffered mode is set then the default multiple writing method write(Pattern [] patterns) method is called to loop through all the patterns calling the single pattern writing abstract method write(Pattern pattern) to write the pattern to the appropriate media. If the synapse is not in buffered mode then the pattern is simply passed to the single pattern writing abstract method write(Pattern pattern) to write the pattern to the appropriate media. Custom xxxOutputSynapse classes can over ride this method if they would like to perform a custom buffering implementation. In this case it will be the responsiblity of the custom class to ensure all patterns are output correctly and any conversion is done prior to writing data.

Specified by:
fwdPut in interface OutputPatternListener
Overrides:
fwdPut in class Synapse
Parameters:
pattern - neural.engine.Pattern

write

public abstract void write(Pattern pattern)
Custom xxxOutputSynapses need to implement at least this method. The custom synapse should write the pattern out to the appropriate media. All patterns including the ending pattern with pattern.getCount()=-1 will be passed to this method so that custom output synapses can perform final processing.


getSeparator

public char getSeparator()
Returns the column separator creation date: (23/04/00 0.50.18)

Returns:
char

revGet

public Pattern revGet()
Description copied from interface: OutputPatternListener
Returns the error pattern coming from the next layer during the training phase

Specified by:
revGet in interface OutputPatternListener
Overrides:
revGet in class Synapse
Returns:
neural.engine.Pattern

setArrays

protected void setArrays(int rows,
                         int cols)
setArrays method comment.

Specified by:
setArrays in class Synapse

setDimensions

protected void setDimensions(int rows,
                             int cols)
Dimensiona l'elemento

Specified by:
setDimensions in class Synapse
Parameters:
int - rows - righe
int - cols - colonne

setSeparator

public void setSeparator(char newSeparator)
Inserire qui la descrizione del metodo. Data di creazione: (23/04/00 0.50.18)

Parameters:
newSeparator - char

setBuffered

public void setBuffered(boolean buf)
Sets the buffered status of this synapse.


isBuffered

public boolean isBuffered()
Checks the buffered status of this synapse.


check

public java.util.TreeSet check()
Description copied from class: Synapse
Base for check messages. Subclasses should call this method from thier own check method.

Specified by:
check in interface NeuralElement
Overrides:
check in class Synapse
Returns:
validation errors.
See Also:
InputPaternListener, OutputPaternListener

getFifo

protected Fifo getFifo()
Getter for property fifo.

Returns:
Value of property fifo.

getPlugIn

public OutputConverterPlugIn getPlugIn()
Returns:
neural.engine.ConverterPlugIn

setPlugIn

public boolean setPlugIn(OutputConverterPlugIn newPlugIn)
Deprecated. use addPlugIn(OutputConverterPlugIn)

Sets the plugin for the data postprocessing

Parameters:
newPlugIn - neural.engine.OutputConverterPlugIn

addPlugIn

public boolean addPlugIn(OutputConverterPlugIn aNewPlugIn)
Adds a plug in to the stream output synapse for data preprocessing. If one or more plug ins are already added to the this synapse, the plug in will be added at the end of the list of plug ins.

Parameters:
aNewPlugIn - The new converter plug in to add (at the end of the list).
Returns:
true when the plug in is added, false when the plug in is not added, e.g. in case the plug in is already added / connected to another synapse.

removeAllPlugIns

public void removeAllPlugIns()
Removes (and disconnects) all (cascading) plug-ins.


dataChanged

public void dataChanged(PlugInEvent data)
Description copied from interface: PlugInListener
This method is called by plug-ins whenever data is changed.

Specified by:
dataChanged in interface PlugInListener
Parameters:
data - the event that is send, i.e. the event indicating that the data is changed.


Submit Feedback to pmarrone@users.sourceforge.net