learner that performs gradient boosting for a specific objective function. It does training and prediction. More...

Inheritance diagram for xgboost::LearnerImpl:

Public Member Functions
	LearnerImpl (std::vector< std::shared_ptr< DMatrix > > cache)

std::vector< std::string >	DumpModel (const FeatureMap &fmap, bool with_stats, std::string format) override
	dump the model in the requested format

Learner *	Slice (bst_layer_t begin, bst_layer_t end, bst_layer_t step, bool *out_of_bound) override
	Slice the model.

void	UpdateOneIter (int iter, std::shared_ptr< DMatrix > train) override
	update the model for one iteration With the specified objective function.

void	BoostOneIter (int iter, std::shared_ptr< DMatrix > train, HostDeviceVector< GradientPair > *in_gpair) override
	Do customized gradient boosting with in_gpair. in_gair can be mutated after this call.

std::string	EvalOneIter (int iter, const std::vector< std::shared_ptr< DMatrix > > &data_sets, const std::vector< std::string > &data_names) override
	evaluate the model for specific iteration using the configured metrics.

void	Predict (std::shared_ptr< DMatrix > data, bool output_margin, HostDeviceVector< bst_float > *out_preds, bst_layer_t layer_begin, bst_layer_t layer_end, bool training, bool pred_leaf, bool pred_contribs, bool approx_contribs, bool pred_interactions) override
	get prediction given the model.

int32_t	BoostedRounds () const override

uint32_t	Groups () const override
	Get the number of output groups from the model.

XGBAPIThreadLocalEntry &	GetThreadLocal () const override

void	InplacePredict (std::shared_ptr< DMatrix > p_m, PredictionType type, float missing, HostDeviceVector< float > **out_preds, bst_layer_t iteration_begin, bst_layer_t iteration_end) override
	Inplace prediction.

void	CalcFeatureScore (std::string const &importance_type, common::Span< int32_t const > trees, std::vector< bst_feature_t > features, std::vector< float > scores) override
	Calculate feature score. See doc in C API for outputs.

const std::map< std::string, std::string > &	GetConfigurationArguments () const override
	Get configuration arguments currently stored by the learner.

Public Member Functions inherited from xgboost::LearnerIO
	LearnerIO (std::vector< std::shared_ptr< DMatrix > > cache)

void	LoadModel (Json const &in) override
	load the model from a JSON object

void	SaveModel (Json *p_out) const override
	saves the model config to a JSON object

void	LoadModel (dmlc::Stream *fi) override

void	SaveModel (dmlc::Stream *fo) const override

void	Save (dmlc::Stream *fo) const override
	saves the model to a stream

void	Load (dmlc::Stream *fi) override
	load the model from a stream

Public Member Functions inherited from xgboost::LearnerConfiguration
	LearnerConfiguration (std::vector< std::shared_ptr< DMatrix > > cache)

void	Configure () override
	Configure Learner based on set parameters.

void	CheckModelInitialized () const

void	LoadConfig (Json const &in) override
	Load configuration from JSON object.

void	SaveConfig (Json *p_out) const override
	Save configuration to JSON object.

void	SetParam (const std::string &key, const std::string &value) override
	Set parameter for booster.

void	SetParams (std::vector< std::pair< std::string, std::string > > const &args) override
	Set multiple parameters at once.

uint32_t	GetNumFeature () const override
	Get the number of features of the booster.

void	SetAttr (const std::string &key, const std::string &value) override
	Set additional attribute to the Booster.

bool	GetAttr (const std::string &key, std::string *out) const override
	Get attribute from the booster. The property will be saved along the booster.

bool	DelAttr (const std::string &key) override
	Delete an attribute from the booster.

void	SetFeatureNames (std::vector< std::string > const &fn) override
	Set the feature names for current booster.

void	GetFeatureNames (std::vector< std::string > *fn) const override
	Get the feature names for current booster.

void	SetFeatureTypes (std::vector< std::string > const &ft) override
	Set the feature types for current booster.

void	GetFeatureTypes (std::vector< std::string > *p_ft) const override
	Get the feature types for current booster.

std::vector< std::string >	GetAttrNames () const override
	Get a vector of attribute names from the booster.

Context const *	Ctx () const override
	Return the context object of this Booster.

Public Member Functions inherited from xgboost::Learner
	~Learner () override
	virtual destructor

Public Member Functions inherited from dmlc::Serializable
virtual	~Serializable ()
	virtual destructor

Protected Member Functions
void	PredictRaw (DMatrix data, PredictionCacheEntry out_preds, bool training, unsigned layer_begin, unsigned layer_end) const
	get un-transformed prediction

void	ValidateDMatrix (DMatrix *p_fmat, bool is_training) const

Protected Member Functions inherited from xgboost::LearnerConfiguration
void	ConfigureModelParamWithoutBaseScore ()

void	InitBaseScore (DMatrix const *p_fmat)
	Calculate the `base_score` based on input data.

Additional Inherited Members
Static Public Member Functions inherited from xgboost::Learner
static Learner *	Create (const std::vector< std::shared_ptr< DMatrix > > &cache_data)
	Create a new instance of learner.

Protected Attributes inherited from xgboost::LearnerConfiguration
std::atomic< bool >	need_configuration_

std::map< std::string, std::string >	cfg_

std::map< std::string, std::string >	attributes_

std::vector< std::string >	feature_names_

std::vector< std::string >	feature_types_

common::Monitor	monitor_

LearnerModelParamLegacy	mparam_

LearnerModelParam	learner_model_param_

LearnerTrainParam	tparam_

PredictionContainer	prediction_container_

std::vector< std::string >	metric_names_

Protected Attributes inherited from xgboost::Learner
std::unique_ptr< ObjFunction >	obj_
	objective function

std::unique_ptr< GradientBooster >	gbm_
	The gradient booster used by the model.

std::vector< std::unique_ptr< Metric > >	metrics_
	The evaluation metrics used to evaluate the model.

Context	ctx_
	Training parameter.

Static Protected Attributes inherited from xgboost::LearnerConfiguration
static std::string const	kEvalMetric {"eval_metric"}

Detailed Description

learner that performs gradient boosting for a specific objective function. It does training and prediction.

Member Function Documentation

◆ BoostedRounds()

int32_t xgboost::LearnerImpl::BoostedRounds ( ) const

inlineoverridevirtual

Implements xgboost::Learner.

◆ BoostOneIter()

void xgboost::LearnerImpl::BoostOneIter	(	int	iter,
		std::shared_ptr< DMatrix >	train,
		HostDeviceVector< GradientPair > *	in_gpair
	)

inlineoverridevirtual

Do customized gradient boosting with in_gpair. in_gair can be mutated after this call.

Parameters

iter	current iteration number
train	reference to the data matrix.
in_gpair	The input gradient statistics.

Implements xgboost::Learner.

◆ CalcFeatureScore()

void xgboost::LearnerImpl::CalcFeatureScore	(	std::string const &	importance_type,
		common::Span< int32_t const >	trees,
		std::vector< bst_feature_t > *	features,
		std::vector< float > *	scores
	)

inlineoverridevirtual

Calculate feature score. See doc in C API for outputs.

Implements xgboost::Learner.

◆ DumpModel()

std::vector< std::string > xgboost::LearnerImpl::DumpModel	(	const FeatureMap &	fmap,
		bool	with_stats,
		std::string	format
	)

inlineoverridevirtual

dump the model in the requested format

Parameters

fmap	feature map that may help give interpretations of feature
with_stats	extra statistics while dumping model
format	the format to dump the model in

Returns: a vector of dump for boosters.

Implements xgboost::Learner.

◆ EvalOneIter()

std::string xgboost::LearnerImpl::EvalOneIter	(	int	iter,
		const std::vector< std::shared_ptr< DMatrix > > &	data_sets,
		const std::vector< std::string > &	data_names
	)

inlineoverridevirtual

evaluate the model for specific iteration using the configured metrics.

Parameters

iter	iteration number
data_sets	datasets to be evaluated.
data_names	name of each dataset

Returns: a string corresponding to the evaluation result

Implements xgboost::Learner.

◆ GetConfigurationArguments()

const std::map< std::string, std::string > & xgboost::LearnerImpl::GetConfigurationArguments ( ) const

inlineoverridevirtual

Get configuration arguments currently stored by the learner.

Returns: Key-value pairs representing configuration arguments

Reimplemented from xgboost::LearnerConfiguration.

◆ GetThreadLocal()

XGBAPIThreadLocalEntry & xgboost::LearnerImpl::GetThreadLocal ( ) const

inlineoverridevirtual

Implements xgboost::Learner.

◆ Groups()

uint32_t xgboost::LearnerImpl::Groups ( ) const

inlineoverridevirtual

Get the number of output groups from the model.

Implements xgboost::Learner.

◆ InplacePredict()

void xgboost::LearnerImpl::InplacePredict	(	std::shared_ptr< DMatrix >	p_m,
		PredictionType	type,
		float	missing,
		HostDeviceVector< float > **	out_preds,
		bst_layer_t	layer_begin,
		bst_layer_t	layer_end
	)

inlineoverridevirtual

Inplace prediction.

Parameters

	p_fmat	A proxy DMatrix that contains the data and related meta info.
	type	Prediction type.
	missing	Missing value in the data.
[in,out]	out_preds	Pointer to output prediction vector.
	layer_begin	Beginning of boosted tree layer used for prediction.
	layer_end	End of booster layer. 0 means do not limit trees.

Implements xgboost::Learner.

◆ Predict()

void xgboost::LearnerImpl::Predict	(	std::shared_ptr< DMatrix >	data,
		bool	output_margin,
		HostDeviceVector< bst_float > *	out_preds,
		bst_layer_t	layer_begin,
		bst_layer_t	layer_end,
		bool	training,
		bool	pred_leaf,
		bool	pred_contribs,
		bool	approx_contribs,
		bool	pred_interactions
	)

inlineoverridevirtual

get prediction given the model.

Parameters

data	input data
output_margin	whether to only predict margin value instead of transformed prediction
out_preds	output vector that stores the prediction
layer_begin	Beginning of boosted tree layer used for prediction.
layer_end	End of booster layer. 0 means do not limit trees.
training	Whether the prediction result is used for training
pred_leaf	whether to only predict the leaf index of each tree in a boosted tree predictor
pred_contribs	whether to only predict the feature contributions
approx_contribs	whether to approximate the feature contributions for speed
pred_interactions	whether to compute the feature pair contributions

Implements xgboost::Learner.

◆ PredictRaw()

void xgboost::LearnerImpl::PredictRaw	(	DMatrix *	data,
		PredictionCacheEntry *	out_preds,
		bool	training,
		unsigned	layer_begin,
		unsigned	layer_end
	)		const

inlineprotected

get un-transformed prediction

Parameters

data	training data matrix
out_preds	output vector that stores the prediction
ntree_limit	limit number of trees used for boosted tree predictor, when it equals 0, this means we are using all the trees
training	allow dropout when the DART booster is being used

◆ Slice()

Learner * xgboost::LearnerImpl::Slice	(	bst_layer_t	begin,
		bst_layer_t	end,
		bst_layer_t	step,
		bool *	out_of_bound
	)

inlineoverridevirtual

Slice the model.

See InplacePredict for layer parameters.

Parameters

step	step size between slice.
out_of_bound	Return true if end layer is out of bound.

Returns: a sliced model.

Implements xgboost::Learner.

◆ UpdateOneIter()

void xgboost::LearnerImpl::UpdateOneIter	(	int	iter,
		std::shared_ptr< DMatrix >	train
	)

inlineoverridevirtual

update the model for one iteration With the specified objective function.

Parameters

iter	current iteration number
train	reference to the data matrix.

Implements xgboost::Learner.

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

External/xgboost/src/learner.cc

Public Member Functions

Protected Member Functions

Additional Inherited Members

Detailed Description

Member Function Documentation

◆ BoostedRounds()

◆ BoostOneIter()

◆ CalcFeatureScore()

◆ DumpModel()

◆ EvalOneIter()

◆ GetConfigurationArguments()

◆ GetThreadLocal()

◆ Groups()

◆ InplacePredict()

◆ Predict()

◆ PredictRaw()

◆ Slice()

◆ UpdateOneIter()