TarzaNN
TarzaNN neural network simulator
Public Member Functions | Public Attributes | Protected Attributes
Monitor Class Reference

#include <Monitor.h>

Inheritance diagram for Monitor:
ODEMonitor

List of all members.

Public Member Functions

 Monitor (Network *n)
virtual ~Monitor ()
void setOutput (vector< OutputLayout > areas)
void setTask ()
void setWTASteps (int steps)
vector< OutputLayout > * getWinners ()
virtual void doneStep ()
void selectWinner (FeaturePlaneAbstract *fp, int selectX, int selectY)
QPoint * getSelectedWinner ()
virtual bool update ()
void resetWinnerPlane ()
int getStep ()
bool getManualOverwrite ()

Public Attributes

FeaturePlaneAbstractwinnerPlane

Protected Attributes

Networknn
vector< OutputLayoutwinningAreas
int step
int wta_steps
bool hasTask
bool manualOverwrite
QPoint * manualWinner

Detailed Description

Monitors the outputs for STM, deciding on top level winners


Constructor & Destructor Documentation

Monitor::Monitor ( Network n)
Monitor::~Monitor ( ) [virtual]

Member Function Documentation

void Monitor::doneStep ( ) [virtual]

Called by the sync mechanism, indicating that one feed forward step has completed, the network is ready for top level competition

Reimplemented in ODEMonitor.

bool Monitor::getManualOverwrite ( )

Has winner been selected manually

QPoint * Monitor::getSelectedWinner ( )

Get manually selected competition winner

int Monitor::getStep ( ) [inline]

Get step count

vector< OutputLayout > * Monitor::getWinners ( )

return the vector of top levels

void Monitor::resetWinnerPlane ( )

Reset the winning FeaturePlane

void Monitor::selectWinner ( FeaturePlaneAbstract fp,
int  selectX,
int  selectY 
)

Manually select competition winner. Called from the UI

void Monitor::setOutput ( vector< OutputLayout areas)

Sets the "top level" - a vector of OutputLayout objects

void Monitor::setTask ( )

Sets the task flag

void Monitor::setWTASteps ( int  steps)

Sets the number of steps the WTA needs

bool Monitor::update ( ) [virtual]

Predicate, should the Feature Planes determine top-level winners

Reimplemented in ODEMonitor.


Member Data Documentation

bool Monitor::hasTask [protected]

Do we have a task

bool Monitor::manualOverwrite [protected]

Has winner been manually selected

QPoint* Monitor::manualWinner [protected]

Manually selected winner

Network* Monitor::nn [protected]

Neural network

int Monitor::step [protected]

Number of steps through the network

The winning FeaturePlane

vector<OutputLayout> Monitor::winningAreas [protected]

The "top level" - a vector of OutputLayout objects

int Monitor::wta_steps [protected]

The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines