#include <boosting.h>

Inheritance diagram for LightGBM::Boosting:

Public Member Functions
virtual	~Boosting ()
	virtual destructor

virtual void	Init (const Config config, const Dataset train_data, const ObjectiveFunction objective_function, const std::vector< const Metric > &training_metrics)=0
	Initialization logic.

virtual void	MergeFrom (const Boosting *other)=0
	Merge model from other boosting object Will insert to the front of current boosting object.

virtual void	ShuffleModels (int start_iter, int end_iter)=0
	Shuffle Existing Models.

virtual void	ResetTrainingData (const Dataset train_data, const ObjectiveFunction objective_function, const std::vector< const Metric * > &training_metrics)=0

virtual void	ResetConfig (const Config *config)=0

virtual void	AddValidDataset (const Dataset valid_data, const std::vector< const Metric > &valid_metrics)=0
	Add a validation data.

virtual void	Train (int snapshot_freq, const std::string &model_output_path)=0

virtual void	RefitTree (const std::vector< std::vector< int > > &tree_leaf_prediction)=0
	Update the tree output by new training data.

virtual bool	TrainOneIter (const score_t gradients, const score_t hessians)=0
	Training logic.

virtual void	RollbackOneIter ()=0
	Rollback one iteration.

virtual int	GetCurrentIteration () const =0
	return current iteration

virtual std::vector< double >	GetEvalAt (int data_idx) const =0
	Get evaluation result at data_idx data.

virtual const double *	GetTrainingScore (int64_t *out_len)=0
	Get current training score.

virtual int64_t	GetNumPredictAt (int data_idx) const =0
	Get prediction result at data_idx data.

virtual void	GetPredictAt (int data_idx, double result, int64_t out_len)=0
	Get prediction result at data_idx data.

virtual int	NumPredictOneRow (int num_iteration, bool is_pred_leaf, bool is_pred_contrib) const =0

virtual void	PredictRaw (const double features, double output, const PredictionEarlyStopInstance *early_stop) const =0
	Prediction for one record, not sigmoid transform.

virtual void	PredictRawByMap (const std::unordered_map< int, double > &features, double output, const PredictionEarlyStopInstance early_stop) const =0

virtual void	Predict (const double features, double output, const PredictionEarlyStopInstance *early_stop) const =0
	Prediction for one record, sigmoid transformation will be used if needed.

virtual void	PredictByMap (const std::unordered_map< int, double > &features, double output, const PredictionEarlyStopInstance early_stop) const =0

virtual void	PredictLeafIndex (const double features, double output) const =0
	Prediction for one record with leaf index.

virtual void	PredictLeafIndexByMap (const std::unordered_map< int, double > &features, double *output) const =0

virtual void	PredictContrib (const double features, double output, const PredictionEarlyStopInstance *early_stop) const =0
	Feature contributions for the model's prediction of one record.

virtual std::string	DumpModel (int start_iteration, int num_iteration) const =0
	Dump model to json format string.

virtual std::string	ModelToIfElse (int num_iteration) const =0
	Translate model to if-else statement.

virtual bool	SaveModelToIfElse (int num_iteration, const char *filename) const =0
	Translate model to if-else statement.

virtual bool	SaveModelToFile (int start_iteration, int num_iterations, const char *filename) const =0
	Save model to file.

std::string	SaveModelToString (int num_iterations)
	Save model to string.

virtual std::string	SaveModelToString (int start_iteration, int num_iterations) const =0

bool	LoadModelFromString (std::string str)
	Restore from a serialized string.

virtual bool	LoadModelFromString (const char *buffer, size_t len)=0

virtual std::vector< double >	FeatureImportance (int num_iteration, int importance_type) const =0
	Calculate feature importances.

virtual int	MaxFeatureIdx () const =0
	Get max feature index of this model.

virtual std::vector< std::string >	FeatureNames () const =0
	Get feature names of this model.

virtual int	LabelIdx () const =0
	Get index of label column.

virtual int	NumberOfTotalModel () const =0
	Get number of weak sub-models.

virtual int	NumModelPerIteration () const =0
	Get number of models per iteration.

virtual int	NumberOfClasses () const =0
	Get number of classes.

virtual bool	NeedAccuratePrediction () const =0
	The prediction should be accurate or not. True will disable early stopping for prediction.

virtual void	InitPredict (int num_iteration, bool is_pred_contrib)=0
	Initial work for the prediction.

virtual const char *	SubModelName () const =0
	Name of submodel.

Boosting &	operator= (const Boosting &)=delete
	Disable copy.

	Boosting (const Boosting &)=delete
	Disable copy.

Static Public Member Functions
static bool	LoadFileToBoosting (Boosting boosting, const char filename)

static Boosting *	CreateBoosting (const std::string &type, const char *filename)
	Create boosting object.

Detailed Description

The interface for Boosting.

Member Function Documentation

◆ AddValidDataset()

virtual void LightGBM::Boosting::AddValidDataset	(	const Dataset *	valid_data,
		const std::vector< const Metric * > &	valid_metrics
	)

pure virtual

Add a validation data.

Parameters

valid_data	Validation data
valid_metrics	Metric for validation data

Implemented in LightGBM::GBDT, and LightGBM::RF.

◆ CreateBoosting()

Boosting * LightGBM::Boosting::CreateBoosting	(	const std::string &	type,
		const char *	filename
	)

static

Create boosting object.

Parameters

type	Type of boosting
format	Format of model
config	config for boosting
filename	name of model file, if existing will continue to train from this model

Returns: The boosting object

◆ DumpModel()

virtual std::string LightGBM::Boosting::DumpModel	(	int	start_iteration,
		int	num_iteration
	)		const

pure virtual

Dump model to json format string.

Parameters

start_iteration	The model will be saved start from
num_iteration	Number of iterations that want to dump, -1 means dump all

Returns: Json format string of model

Implemented in LightGBM::GBDT.

◆ FeatureImportance()

virtual std::vector< double > LightGBM::Boosting::FeatureImportance	(	int	num_iteration,
		int	importance_type
	)		const

pure virtual

Calculate feature importances.

Parameters

num_iteration	Number of model that want to use for feature importance, -1 means use all
importance_type	0 for split, 1 for gain

Returns: vector of feature_importance

Implemented in LightGBM::GBDT.

◆ FeatureNames()

virtual std::vector< std::string > LightGBM::Boosting::FeatureNames ( ) const

pure virtual

Get feature names of this model.

Returns: Feature names of this model

Implemented in LightGBM::GBDT.

◆ GetCurrentIteration()

virtual int LightGBM::Boosting::GetCurrentIteration ( ) const

pure virtual

return current iteration

Implemented in LightGBM::GBDT.

◆ GetEvalAt()

virtual std::vector< double > LightGBM::Boosting::GetEvalAt ( int data_idx ) const

pure virtual

Get evaluation result at data_idx data.

Parameters

data_idx 0: training data, 1: 1st validation data

Returns: evaluation result

Implemented in LightGBM::GBDT.

◆ GetNumPredictAt()

virtual int64_t LightGBM::Boosting::GetNumPredictAt ( int data_idx ) const

pure virtual

Get prediction result at data_idx data.

Parameters

data_idx 0: training data, 1: 1st validation data

Returns: out_len length of returned score

Implemented in LightGBM::GBDT.

◆ GetPredictAt()

virtual void LightGBM::Boosting::GetPredictAt	(	int	data_idx,
		double *	result,
		int64_t *	out_len
	)

pure virtual

Get prediction result at data_idx data.

Parameters

data_idx	0: training data, 1: 1st validation data
result	used to store prediction result, should allocate memory before call this function
out_len	length of returned score

Implemented in LightGBM::GBDT.

◆ GetTrainingScore()

virtual const double * LightGBM::Boosting::GetTrainingScore ( int64_t * out_len )

pure virtual

Get current training score.

Parameters

out_len length of returned score

Returns: training score

Implemented in LightGBM::DART, and LightGBM::GBDT.

◆ Init()

virtual void LightGBM::Boosting::Init	(	const Config *	config,
		const Dataset *	train_data,
		const ObjectiveFunction *	objective_function,
		const std::vector< const Metric * > &	training_metrics
	)

pure virtual

Initialization logic.

Parameters

config	Configs for boosting
train_data	Training data
objective_function	Training objective function
training_metrics	Training metric

Implemented in LightGBM::DART, LightGBM::GOSS, LightGBM::RF, and LightGBM::GBDT.

◆ InitPredict()

virtual void LightGBM::Boosting::InitPredict	(	int	num_iteration,
		bool	is_pred_contrib
	)

pure virtual

Initial work for the prediction.

Parameters

num_iteration	number of used iteration
is_pred_contrib

Implemented in LightGBM::GBDT.

◆ LabelIdx()

virtual int LightGBM::Boosting::LabelIdx ( ) const

pure virtual

Get index of label column.

Returns: index of label column

Implemented in LightGBM::GBDT.

◆ LoadModelFromString()

bool LightGBM::Boosting::LoadModelFromString ( std::string str )

inline

Restore from a serialized string.

Parameters

buffer	The content of model
len	The length of buffer

Returns: true if succeeded

◆ MaxFeatureIdx()

virtual int LightGBM::Boosting::MaxFeatureIdx ( ) const

pure virtual

Get max feature index of this model.

Returns: Max feature index of this model

Implemented in LightGBM::GBDT.

◆ MergeFrom()

virtual void LightGBM::Boosting::MergeFrom ( const Boosting * other )

pure virtual

Merge model from other boosting object Will insert to the front of current boosting object.

Parameters

other

Implemented in LightGBM::GBDT.

◆ ModelToIfElse()

virtual std::string LightGBM::Boosting::ModelToIfElse ( int num_iteration ) const

pure virtual

Translate model to if-else statement.

Parameters

num_iteration Number of iterations that want to translate, -1 means translate all

Returns: if-else format codes of model

Implemented in LightGBM::GBDT.

◆ NeedAccuratePrediction()

virtual bool LightGBM::Boosting::NeedAccuratePrediction ( ) const

pure virtual

The prediction should be accurate or not. True will disable early stopping for prediction.

Implemented in LightGBM::GBDT, and LightGBM::RF.

◆ NumberOfClasses()

virtual int LightGBM::Boosting::NumberOfClasses ( ) const

pure virtual

Get number of classes.

Returns: Number of classes

Implemented in LightGBM::GBDT.

◆ NumberOfTotalModel()

virtual int LightGBM::Boosting::NumberOfTotalModel ( ) const

pure virtual

Get number of weak sub-models.

Returns: Number of weak sub-models

Implemented in LightGBM::GBDT.

◆ NumModelPerIteration()

virtual int LightGBM::Boosting::NumModelPerIteration ( ) const

pure virtual

Get number of models per iteration.

Returns: Number of models per iteration

Implemented in LightGBM::GBDT.

◆ NumPredictOneRow()

virtual int LightGBM::Boosting::NumPredictOneRow	(	int	num_iteration,
		bool	is_pred_leaf,
		bool	is_pred_contrib
	)		const

pure virtual

Implemented in LightGBM::GBDT.

◆ Predict()

virtual void LightGBM::Boosting::Predict	(	const double *	features,
		double *	output,
		const PredictionEarlyStopInstance *	early_stop
	)		const

pure virtual

Prediction for one record, sigmoid transformation will be used if needed.

Parameters

feature_values	Feature value on this record
output	Prediction result for this record
early_stop	Early stopping instance. If nullptr, no early stopping is applied and all models are evaluated.

Implemented in LightGBM::GBDT.

◆ PredictContrib()

virtual void LightGBM::Boosting::PredictContrib	(	const double *	features,
		double *	output,
		const PredictionEarlyStopInstance *	early_stop
	)		const

pure virtual

Feature contributions for the model's prediction of one record.

Parameters

feature_values	Feature value on this record
output	Prediction result for this record
early_stop	Early stopping instance. If nullptr, no early stopping is applied and all models are evaluated.

Implemented in LightGBM::GBDT.

◆ PredictLeafIndex()

virtual void LightGBM::Boosting::PredictLeafIndex	(	const double *	features,
		double *	output
	)		const

pure virtual

Prediction for one record with leaf index.

Parameters

feature_values	Feature value on this record
output	Prediction result for this record

Implemented in LightGBM::GBDT.

◆ PredictRaw()

virtual void LightGBM::Boosting::PredictRaw	(	const double *	features,
		double *	output,
		const PredictionEarlyStopInstance *	early_stop
	)		const

pure virtual

Prediction for one record, not sigmoid transform.

Parameters

feature_values	Feature value on this record
output	Prediction result for this record
early_stop	Early stopping instance. If nullptr, no early stopping is applied and all models are evaluated.

Implemented in LightGBM::GBDT.

◆ RefitTree()

virtual void LightGBM::Boosting::RefitTree ( const std::vector< std::vector< int > > & tree_leaf_prediction )

pure virtual

Update the tree output by new training data.

Implemented in LightGBM::GBDT.

◆ ResetConfig()

virtual void LightGBM::Boosting::ResetConfig ( const Config * config )

pure virtual

Implemented in LightGBM::GOSS, LightGBM::RF, and LightGBM::GBDT.

◆ ResetTrainingData()

virtual void LightGBM::Boosting::ResetTrainingData	(	const Dataset *	train_data,
		const ObjectiveFunction *	objective_function,
		const std::vector< const Metric * > &	training_metrics
	)

pure virtual

Implemented in LightGBM::GBDT, LightGBM::GOSS, and LightGBM::RF.

◆ RollbackOneIter()

virtual void LightGBM::Boosting::RollbackOneIter ( )

pure virtual

Rollback one iteration.

Implemented in LightGBM::GBDT, and LightGBM::RF.

◆ SaveModelToFile()

virtual bool LightGBM::Boosting::SaveModelToFile	(	int	start_iteration,
		int	num_iterations,
		const char *	filename
	)		const

pure virtual

Save model to file.

Parameters

start_iteration	The model will be saved start from
num_iterations	Number of model that want to save, -1 means save all
is_finish	Is training finished or not
filename	Filename that want to save to

Returns: true if succeeded

Implemented in LightGBM::GBDT.

◆ SaveModelToIfElse()

virtual bool LightGBM::Boosting::SaveModelToIfElse	(	int	num_iteration,
		const char *	filename
	)		const

pure virtual

Translate model to if-else statement.

Parameters

num_iteration	Number of iterations that want to translate, -1 means translate all
filename	Filename that want to save to

Returns: is_finish Is training finished or not

Implemented in LightGBM::GBDT.

◆ SaveModelToString()

std::string LightGBM::Boosting::SaveModelToString ( int num_iterations )

inline

Save model to string.

Parameters

start_iteration	The model will be saved start from
num_iterations	Number of model that want to save, -1 means save all

Returns: Non-empty string if succeeded

◆ ShuffleModels()

virtual void LightGBM::Boosting::ShuffleModels	(	int	start_iter,
		int	end_iter
	)

pure virtual

Shuffle Existing Models.

Implemented in LightGBM::GBDT.

◆ SubModelName()

virtual const char * LightGBM::Boosting::SubModelName ( ) const

pure virtual

Name of submodel.

Implemented in LightGBM::GBDT.

◆ Train()

virtual void LightGBM::Boosting::Train	(	int	snapshot_freq,
		const std::string &	model_output_path
	)

pure virtual

Implemented in LightGBM::GBDT.

◆ TrainOneIter()

virtual bool LightGBM::Boosting::TrainOneIter	(	const score_t *	gradients,
		const score_t *	hessians
	)

pure virtual

Training logic.

Parameters

gradients	nullptr for using default objective, otherwise use self-defined boosting
hessians	nullptr for using default objective, otherwise use self-defined boosting

Returns: True if cannot train anymore

Implemented in LightGBM::DART, LightGBM::GBDT, and LightGBM::RF.

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

External/LightGBM_2.2.3/LightGBM-2.2.3/include/LightGBM/boosting.h
External/LightGBM_2.2.3/LightGBM-2.2.3/src/boosting/boosting.cpp

Public Member Functions

Static Public Member Functions

Detailed Description

Member Function Documentation

◆ AddValidDataset()

◆ CreateBoosting()

◆ DumpModel()

◆ FeatureImportance()

◆ FeatureNames()

◆ GetCurrentIteration()

◆ GetEvalAt()

◆ GetNumPredictAt()

◆ GetPredictAt()

◆ GetTrainingScore()

◆ Init()

◆ InitPredict()

◆ LabelIdx()

◆ LoadModelFromString()

◆ MaxFeatureIdx()

◆ MergeFrom()

◆ ModelToIfElse()

◆ NeedAccuratePrediction()

◆ NumberOfClasses()

◆ NumberOfTotalModel()

◆ NumModelPerIteration()

◆ NumPredictOneRow()

◆ Predict()

◆ PredictContrib()

◆ PredictLeafIndex()

◆ PredictRaw()

◆ RefitTree()

◆ ResetConfig()

◆ ResetTrainingData()

◆ RollbackOneIter()

◆ SaveModelToFile()

◆ SaveModelToIfElse()

◆ SaveModelToString()

◆ ShuffleModels()

◆ SubModelName()

◆ Train()

◆ TrainOneIter()