#include <opennn/evaluation/evaluation.h>

Classes
struct	BinaryClassificationRates

struct	GoodnessOfFitAnalysis

struct	ReconstructionErrorStatistics

struct	RocAnalysis

Public Member Functions
	Evaluation (Network =nullptr, Dataset =nullptr)

const Network *	get_network () const

const Dataset *	get_dataset () const

void	set_network (Network *new_network)

void	set_dataset (Dataset *new_dataset)

void	set_batch_size (Index new_batch_size)

void	check () const

pair< MatrixR, MatrixR >	get_targets_and_outputs (const string &sample_role) const

Tensor3	calculate_error_data () const

MatrixR	calculate_percentage_error_data () const

vector< vector< Descriptives > >	calculate_error_data_descriptives () const

vector< Histogram >	calculate_error_data_histograms (const Index=10) const

VectorR	calculate_errors (const MatrixR &, const MatrixR &) const

VectorR	calculate_errors (const string &) const

float	calculate_determination (const VectorR &, const VectorR &) const

Tensor< GoodnessOfFitAnalysis, 1 >	perform_goodness_of_fit_analysis () const

void	print_goodness_of_fit_analysis () const

VectorR	calculate_reconstruction_errors (const MatrixR &, const MatrixR &) const

VectorR	calculate_reconstruction_errors (const string &) const

ReconstructionErrorStatistics	calculate_reconstruction_error_statistics (const VectorR &) const

float	calculate_anomaly_threshold (const ReconstructionErrorStatistics &, float=1.0f) const

VectorI	calculate_anomaly_predictions (const VectorR &, float) const

VectorR	calculate_binary_classification_errors (const string &sample_role) const

VectorR	calculate_multiple_classification_errors (const string &sample_role) const

VectorR	calculate_binary_classification_tests (const float=0.50) const

VectorR	calculate_binary_classification_tests (const MatrixR &, const MatrixR &, float=0.50) const

void	print_binary_classification_tests () const

void	print_multiple_classification_tests () const

MatrixI	calculate_confusion (const MatrixR &, const MatrixR &, float=0.50) const

MatrixI	calculate_confusion (const float=0.50) const

VectorI	calculate_positives_negatives_rate (const MatrixR &, const MatrixR &) const

BinaryClassificationRates	calculate_binary_classification_rates (const float=0.50) const

vector< Index >	calculate_true_positive_samples (const MatrixR &targets, const MatrixR &outputs, const vector< Index > &indices, float threshold) const

vector< Index >	calculate_false_positive_samples (const MatrixR &targets, const MatrixR &outputs, const vector< Index > &indices, float threshold) const

vector< Index >	calculate_false_negative_samples (const MatrixR &targets, const MatrixR &outputs, const vector< Index > &indices, float threshold) const

vector< Index >	calculate_true_negative_samples (const MatrixR &targets, const MatrixR &outputs, const vector< Index > &indices, float threshold) const

Tensor< VectorI, 2 >	calculate_multiple_classification_rates () const

Tensor< VectorI, 2 >	calculate_multiple_classification_rates (const MatrixR &, const MatrixR &, const vector< Index > &) const

RocAnalysis	perform_roc_analysis () const

MatrixR	calculate_roc_curve (const MatrixR &, const MatrixR &) const

float	calculate_area_under_curve (const MatrixR &) const

float	calculate_area_under_curve_confidence_limit (const MatrixR &, const MatrixR &) const

float	calculate_area_under_curve_confidence_limit (float, Index, Index) const

float	calculate_optimal_threshold (const MatrixR &) const

MatrixR	perform_lift_chart_analysis () const

MatrixR	calculate_cumulative_gain (const MatrixR &, const MatrixR &) const

MatrixR	calculate_lift_chart (const MatrixR &) const

Constructor & Destructor Documentation

◆ Evaluation()

opennn::Evaluation::Evaluation	(	Network *	= nullptr,
		Dataset *	= nullptr )

explicit

Member Function Documentation

◆ calculate_anomaly_predictions()

VectorI opennn::Evaluation::calculate_anomaly_predictions	(	const VectorR &	,
		float	) const

◆ calculate_anomaly_threshold()

float opennn::Evaluation::calculate_anomaly_threshold	(	const ReconstructionErrorStatistics &	,
		float	= 1.0f ) const

◆ calculate_area_under_curve()

float opennn::Evaluation::calculate_area_under_curve ( const MatrixR & ) const

◆ calculate_area_under_curve_confidence_limit() [1/2]

float opennn::Evaluation::calculate_area_under_curve_confidence_limit	(	const MatrixR &	,
		const MatrixR &	) const

◆ calculate_area_under_curve_confidence_limit() [2/2]

float opennn::Evaluation::calculate_area_under_curve_confidence_limit	(	float	,
		Index	,
		Index	) const

◆ calculate_binary_classification_errors()

VectorR opennn::Evaluation::calculate_binary_classification_errors ( const string & sample_role ) const

inline

◆ calculate_binary_classification_rates()

BinaryClassificationRates opennn::Evaluation::calculate_binary_classification_rates ( const float = 0.50 ) const

◆ calculate_binary_classification_tests() [1/2]

VectorR opennn::Evaluation::calculate_binary_classification_tests ( const float = 0.50 ) const

◆ calculate_binary_classification_tests() [2/2]

VectorR opennn::Evaluation::calculate_binary_classification_tests	(	const MatrixR &	,
		const MatrixR &	,
		float	= 0.50 ) const

◆ calculate_confusion() [1/2]

MatrixI opennn::Evaluation::calculate_confusion ( const float = 0.50 ) const

◆ calculate_confusion() [2/2]

MatrixI opennn::Evaluation::calculate_confusion	(	const MatrixR &	,
		const MatrixR &	,
		float	= 0.50 ) const

◆ calculate_cumulative_gain()

MatrixR opennn::Evaluation::calculate_cumulative_gain	(	const MatrixR &	,
		const MatrixR &	) const

◆ calculate_determination()

float opennn::Evaluation::calculate_determination	(	const VectorR &	,
		const VectorR &	) const

◆ calculate_error_data()

Tensor3 opennn::Evaluation::calculate_error_data ( ) const

◆ calculate_error_data_descriptives()

vector< vector< Descriptives > > opennn::Evaluation::calculate_error_data_descriptives ( ) const

◆ calculate_error_data_histograms()

vector< Histogram > opennn::Evaluation::calculate_error_data_histograms ( const Index = 10 ) const

◆ calculate_errors() [1/2]

VectorR opennn::Evaluation::calculate_errors	(	const MatrixR &	,
		const MatrixR &	) const

◆ calculate_errors() [2/2]

VectorR opennn::Evaluation::calculate_errors ( const string & ) const

◆ calculate_false_negative_samples()

vector< Index > opennn::Evaluation::calculate_false_negative_samples	(	const MatrixR &	targets,
		const MatrixR &	outputs,
		const vector< Index > &	indices,
		float	threshold ) const

inline

◆ calculate_false_positive_samples()

vector< Index > opennn::Evaluation::calculate_false_positive_samples	(	const MatrixR &	targets,
		const MatrixR &	outputs,
		const vector< Index > &	indices,
		float	threshold ) const

inline

◆ calculate_lift_chart()

MatrixR opennn::Evaluation::calculate_lift_chart ( const MatrixR & ) const

◆ calculate_multiple_classification_errors()

VectorR opennn::Evaluation::calculate_multiple_classification_errors ( const string & sample_role ) const

inline

◆ calculate_multiple_classification_rates() [1/2]

Tensor< VectorI, 2 > opennn::Evaluation::calculate_multiple_classification_rates ( ) const

◆ calculate_multiple_classification_rates() [2/2]

Tensor< VectorI, 2 > opennn::Evaluation::calculate_multiple_classification_rates	(	const MatrixR &	,
		const MatrixR &	,
		const vector< Index > &	) const

◆ calculate_optimal_threshold()

float opennn::Evaluation::calculate_optimal_threshold ( const MatrixR & ) const

◆ calculate_percentage_error_data()

MatrixR opennn::Evaluation::calculate_percentage_error_data ( ) const

◆ calculate_positives_negatives_rate()

VectorI opennn::Evaluation::calculate_positives_negatives_rate	(	const MatrixR &	,
		const MatrixR &	) const

◆ calculate_reconstruction_error_statistics()

ReconstructionErrorStatistics opennn::Evaluation::calculate_reconstruction_error_statistics ( const VectorR & ) const

◆ calculate_reconstruction_errors() [1/2]

VectorR opennn::Evaluation::calculate_reconstruction_errors	(	const MatrixR &	,
		const MatrixR &	) const

◆ calculate_reconstruction_errors() [2/2]

VectorR opennn::Evaluation::calculate_reconstruction_errors ( const string & ) const

◆ calculate_roc_curve()

MatrixR opennn::Evaluation::calculate_roc_curve	(	const MatrixR &	,
		const MatrixR &	) const

◆ calculate_true_negative_samples()

vector< Index > opennn::Evaluation::calculate_true_negative_samples	(	const MatrixR &	targets,
		const MatrixR &	outputs,
		const vector< Index > &	indices,
		float	threshold ) const

inline

◆ calculate_true_positive_samples()

vector< Index > opennn::Evaluation::calculate_true_positive_samples	(	const MatrixR &	targets,
		const MatrixR &	outputs,
		const vector< Index > &	indices,
		float	threshold ) const

inline

◆ check()

void opennn::Evaluation::check ( ) const

◆ get_dataset()

const Dataset * opennn::Evaluation::get_dataset ( ) const

inline

◆ get_network()

const Network * opennn::Evaluation::get_network ( ) const

inline

◆ get_targets_and_outputs()

pair< MatrixR, MatrixR > opennn::Evaluation::get_targets_and_outputs ( const string & sample_role ) const

◆ perform_goodness_of_fit_analysis()

Tensor< GoodnessOfFitAnalysis, 1 > opennn::Evaluation::perform_goodness_of_fit_analysis ( ) const

◆ perform_lift_chart_analysis()

MatrixR opennn::Evaluation::perform_lift_chart_analysis ( ) const

◆ perform_roc_analysis()

RocAnalysis opennn::Evaluation::perform_roc_analysis ( ) const

◆ print_binary_classification_tests()

void opennn::Evaluation::print_binary_classification_tests ( ) const

◆ print_goodness_of_fit_analysis()

void opennn::Evaluation::print_goodness_of_fit_analysis ( ) const

◆ print_multiple_classification_tests()

void opennn::Evaluation::print_multiple_classification_tests ( ) const

◆ set_batch_size()

void opennn::Evaluation::set_batch_size ( Index new_batch_size )

inline

◆ set_dataset()

void opennn::Evaluation::set_dataset ( Dataset * new_dataset )

inline

◆ set_network()

void opennn::Evaluation::set_network ( Network * new_network )

inline

Classes

Public Member Functions

Constructor & Destructor Documentation

◆ Evaluation()

Member Function Documentation

◆ calculate_anomaly_predictions()

◆ calculate_anomaly_threshold()

◆ calculate_area_under_curve()

◆ calculate_area_under_curve_confidence_limit() [1/2]

◆ calculate_area_under_curve_confidence_limit() [2/2]

◆ calculate_binary_classification_errors()

◆ calculate_binary_classification_rates()

◆ calculate_binary_classification_tests() [1/2]

◆ calculate_binary_classification_tests() [2/2]

◆ calculate_confusion() [1/2]

◆ calculate_confusion() [2/2]

◆ calculate_cumulative_gain()

◆ calculate_determination()

◆ calculate_error_data()

◆ calculate_error_data_descriptives()

◆ calculate_error_data_histograms()

◆ calculate_errors() [1/2]

◆ calculate_errors() [2/2]

◆ calculate_false_negative_samples()

◆ calculate_false_positive_samples()

◆ calculate_lift_chart()

◆ calculate_multiple_classification_errors()

◆ calculate_multiple_classification_rates() [1/2]

◆ calculate_multiple_classification_rates() [2/2]

◆ calculate_optimal_threshold()

◆ calculate_percentage_error_data()

◆ calculate_positives_negatives_rate()

◆ calculate_reconstruction_error_statistics()

◆ calculate_reconstruction_errors() [1/2]

◆ calculate_reconstruction_errors() [2/2]

◆ calculate_roc_curve()

◆ calculate_true_negative_samples()

◆ calculate_true_positive_samples()

◆ check()

◆ get_dataset()

◆ get_network()

◆ get_targets_and_outputs()

◆ perform_goodness_of_fit_analysis()

◆ perform_lift_chart_analysis()

◆ perform_roc_analysis()

◆ print_binary_classification_tests()

◆ print_goodness_of_fit_analysis()

◆ print_multiple_classification_tests()

◆ set_batch_size()

◆ set_dataset()

◆ set_network()