Dataset Class Reference

#include <Dataset.h>

Inheritance diagram for Dataset:

List of all members.

Public Member Functions
	Dataset (char *basefilename)
virtual	~Dataset ()
virtual OutputImage *	getImage (int index, OutputImage **list, int LorR=0)=0
virtual Matrix *	getDesiredOutcome (int index)=0
virtual int	getRandomIndex ()
void	setMaxTestSteps (int mTS)
	Max number of manipulations during testing.
int	getMaxTestSteps ()
	Max number of manipulations during testing.
void	setMaxSteps (int mS)
	How many images to generate based on the same stimulus.
void	setInvarianceLevel (int levels)
	How many images to generate based on the same stimulus.
int	getInvarianceLevel ()
int	getMaxSteps ()
	How many images to generate based on the same stimulus.
bool	isSameAsLast ()
Protected Attributes
int	count
int	maxTestSteps
	Max number of manipulations during testing.
int	maxSteps
	How many images to generate based on the same stimulus.
int	currentStep
	counter [0-maxSteps)
bool	sameAsLast
	Is same category as last stimulus.
int	invarianceLevel
	Level of learning layer - determines level of invariance.

---

Detailed Description

Superclass for learning datasets. A dataset provides images and the associated desired network output for supervised learning.

---

Constructor & Destructor Documentation

Dataset::Dataset

(

char *

basefilename

)

Constructor. Base file name for input images

Dataset::~Dataset

(

)

[virtual]

---

Member Function Documentation

virtual Matrix* Dataset::getDesiredOutcome

(

int

index

)

[pure virtual]

Get category for specific image

Implemented in NORB_Dataset, Caucasian_Dataset, COIL100_Dataset, SOM_GeomDataset, Greeble_Dataset, SOM_Dataset, Faces_Dataset, Rooz_Dataset, Rooz_DatasetOR2, VanRullen_Dataset, Rooz_Dataset5, and Rooz_Dataset2.

virtual OutputImage* Dataset::getImage	(	int	index,
		OutputImage **	list,
		int	LorR = 0
	)		[pure virtual]

Get specific image, Left or Right is 0 or 1

Implemented in NORB_Dataset, Caucasian_Dataset, COIL100_Dataset, SOM_GeomDataset, Greeble_Dataset, SOM_Dataset, Faces_Dataset, Rooz_Dataset, Rooz_DatasetOR2, VanRullen_Dataset, Rooz_Dataset5, and Rooz_Dataset2.

int Dataset::getInvarianceLevel

(

)

[inline]

int Dataset::getMaxSteps

(

)

[inline]

How many images to generate based on the same stimulus.

int Dataset::getMaxTestSteps

(

)

[inline]

Max number of manipulations during testing.

int Dataset::getRandomIndex

(

)

[virtual]

Get random index within legal range [0, count)

Reimplemented in Caucasian_Dataset, COIL100_Dataset, SOM_GeomDataset, and SOM_Dataset.

bool Dataset::isSameAsLast

(

)

[inline]

void Dataset::setInvarianceLevel

(

int

levels

)

[inline]

How many images to generate based on the same stimulus.

void Dataset::setMaxSteps

(

int

mS

)

[inline]

How many images to generate based on the same stimulus.

void Dataset::setMaxTestSteps

(

int

mTS

)

[inline]

Max number of manipulations during testing.

---

Member Data Documentation

int Dataset::count [protected]

int Dataset::currentStep [protected]

counter [0-maxSteps)

int Dataset::invarianceLevel [protected]

Level of learning layer - determines level of invariance.

int Dataset::maxSteps [protected]

How many images to generate based on the same stimulus.

int Dataset::maxTestSteps [protected]

Max number of manipulations during testing.

bool Dataset::sameAsLast [protected]

Is same category as last stimulus.

---

The documentation for this class was generated from the following files:

• C:/Users/albertlr/projects/TarzaNN/TarzaNN/Dataset.h
• C:/Users/albertlr/projects/TarzaNN/TarzaNN/Dataset.cpp