oevislib_net.DataClassification.Dataset Class Reference

Dataset containing features, classes and labels useful for classification purpose. More...

Inheritance diagram for oevislib_net.DataClassification.Dataset:

Public Member Functions
	Dataset ()
	Constructs an empty Dataset object.
	Dataset (Array< Array< float > > inFeatures, Array< string > inHeader)
	Constructs a new Dataset object given input features.
	Dataset (Array< Array< float > > inFeatures, Array< int > inClasses, Array< string > inHeader)
	Constructs a new Dataset object given input features and their corresponding classes.
	Dataset (Array< Array< float > > inFeatures, Array< string > inLabels, Array< string > inHeader)
	Constructs a new Dataset object given input features and their corresponding labels.
Dataset	Copy ()
Array< Array< float > >	GetFeatures ()
	Returns all the features.
Array< int >	GetClasses ()
	Returns all the classes.
Array< string >	GetLabels ()
	Returns all the labels.
Array< string >	GetHeader ()
	Returns the file header.
int	GetSize ()
	Returns the number of samples in the entire dataset.
int	GetLabelsSize ()
	Returns the number of labels in the dataset.
void	SaveToCSV (string inFilename, char inDelimiter, bool inWriteHeader, bool inWriteLabels, int inDecimalPrecision)
	Saves the dataset to a CSV file.
void	Split (float inTrainingRatio, bool inShuffle, Dataset outTrainDataset, Dataset outTestDataset)
	Splits the dataset into training and testing sets.
void	Standardize ()
	Standardize dataset features (i.e., zero mean, unitary standard deviation)
override object	Clone ()
override void	Dispose ()

Static Public Member Functions
static Dataset	LoadFromCSV (string inFilename, int inHeaderLinesToSkip=0, char inDelimiter=',', Optional< int > inHeaderLineIndex=null, Optional< int > inClassIndex=null, Optional< Range > inFeaturesIndices=null)
	Loads a dataset from a CSV file.

Detailed Description

Dataset containing features, classes and labels useful for classification purpose.

It can also manage the splitting into training and testing sets.

Constructor & Destructor Documentation

Dataset() [1/4]

oevislib_net.DataClassification.Dataset.Dataset ( )

inline

Constructs an empty Dataset object.

Dataset() [2/4]

oevislib_net.DataClassification.Dataset.Dataset ( Array< Array< float > > inFeatures, Array< string > inHeader )

inline

Constructs a new Dataset object given input features.

Parameters

inFeatures	Input features.
inHeader	Input header (containing the features' names). If empty, is not considered.

Dataset() [3/4]

oevislib_net.DataClassification.Dataset.Dataset ( Array< Array< float > > inFeatures, Array< int > inClasses, Array< string > inHeader )

inline

Constructs a new Dataset object given input features and their corresponding classes.

Parameters

inFeatures	Input features.
inClasses	Input classes.
inHeader	Input header (containing the features' names). If empty, is not considered.

Dataset() [4/4]

oevislib_net.DataClassification.Dataset.Dataset ( Array< Array< float > > inFeatures, Array< string > inLabels, Array< string > inHeader )

inline

Constructs a new Dataset object given input features and their corresponding labels.

Parameters

inFeatures	Input features.
inLabels	Input labels.
inHeader	Input header (containing the features' names). If empty, is not considered.

Member Function Documentation

GetClasses()

Array< int > oevislib_net.DataClassification.Dataset.GetClasses ( )

inline

Returns all the classes.

Returns

Classes.

GetFeatures()

Array< Array< float > > oevislib_net.DataClassification.Dataset.GetFeatures ( )

inline

Returns all the features.

Returns

Features.

GetHeader()

Array< string > oevislib_net.DataClassification.Dataset.GetHeader ( )

inline

Returns the file header.

Returns

Header.

GetLabels()

Array< string > oevislib_net.DataClassification.Dataset.GetLabels ( )

inline

Returns all the labels.

Returns

Labels.

GetLabelsSize()

int oevislib_net.DataClassification.Dataset.GetLabelsSize ( )

inline

Returns the number of labels in the dataset.

Returns

Labels size.

GetSize()

int oevislib_net.DataClassification.Dataset.GetSize ( )

inline

Returns the number of samples in the entire dataset.

Returns

Dataset size.

LoadFromCSV()

Dataset oevislib_net.DataClassification.Dataset.LoadFromCSV ( string inFilename, int inHeaderLinesToSkip = 0, char inDelimiter = ',', Optional< int > inHeaderLineIndex = null, Optional< int > inClassIndex = null, Optional< Range > inFeaturesIndices = null )

inlinestatic

Loads a dataset from a CSV file.

Parameters

inFilename	Filename.
inHeaderLinesToSkip	Number of header lines (to exclude when extracting features).
inDelimiter	Delimiter between features.
inHeaderLineIndex	Index of the header line (starting from 0 and below the number of inHeaderLinesToSkip). When loaded, the name corresponding to the class (if present), is excluded.
inClassIndex	Index of the class column (starting from 0). If Null, classes are not loaded.
inFeaturesIndices	Indices of the features columns. If Null, all features are loaded.

Returns

Dataset

SaveToCSV()

void oevislib_net.DataClassification.Dataset.SaveToCSV ( string inFilename, char inDelimiter, bool inWriteHeader, bool inWriteLabels, int inDecimalPrecision )

inline

Saves the dataset to a CSV file.

Parameters

inFilename	Filename.
inDelimiter	Delimiter between features.
inWriteHeader	Whether to write the header as the first line.
inWriteLabels	Whether to write the labels in the first column.
inDecimalPrecision	How many decimal places to write.

Split()

void oevislib_net.DataClassification.Dataset.Split ( float inTrainingRatio, bool inShuffle, Dataset outTrainDataset, Dataset outTestDataset )

inline

Splits the dataset into training and testing sets.

Parameters

inTrainingRatio	Ratio of the training set.
inShuffle	Whether to shuffle the dataset before splitting.
outTrainDataset	Training dataset.
outTestDataset	Testing dataset.

Standardize()

void oevislib_net.DataClassification.Dataset.Standardize ( )

inline

Standardize dataset features (i.e., zero mean, unitary standard deviation)