#include <toolkits/supervised_learning/linear_regression_opt_interface.hpp>

Public Member Functions
	linear_regression_opt_interface (const ml_data &_data, const ml_data &_valid_data, linear_regression &_model, bool _feature_rescaling=true)

	~linear_regression_opt_interface ()

void	set_threads (size_t _n_threads)

void	init_feature_rescaling ()

void	rescale_solution (DenseVector &coefs)

size_t	num_examples () const

size_t	num_variables () const

size_t	num_validation_examples () const

std::vector< std::pair< std::string, size_t > >	get_status_header (const std::vector< std::string > &stat_names)

std::vector< std::string >	get_status (const DenseVector &coefs, const std::vector< std::string > &stats)

void	compute_first_order_statistics (const DenseVector &point, DenseVector &gradient, double &function_value, const size_t mbStart=0, const size_t mbSize=-1)

void	compute_second_order_statistics (const DenseVector &point, DenseMatrix &hessian, DenseVector &gradient, double &function_value)

void	compute_validation_first_order_statistics (const DenseVector &point, DenseVector &gradient, double &function_value)

virtual void	compute_hessian (const DenseVector &point, DenseMatrix &hessian)

virtual double	compute_function_value (const DenseVector &point, const size_t mbStart=0, const size_t mbSize=-1)

virtual void	compute_gradient (const DenseVector &point, DenseVector &gradient, const size_t mbStart=0, const size_t mbSize=-1)

virtual void	reset (int seed)

Protected Attributes
size_t	examples = 0

size_t	features = 0

size_t	variables = 0

std::shared_ptr< l2_rescaling >	scaler

bool	feature_rescaling = false

bool	is_dense = false

Detailed Description

Solver interface for the linear regression problem.

Definition at line 37 of file linear_regression_opt_interface.hpp.

Constructor & Destructor Documentation

◆ linear_regression_opt_interface()

turi::supervised::linear_regression_opt_interface::linear_regression_opt_interface	(	const ml_data &	_data,
		const ml_data &	_valid_data,
		linear_regression &	_model,
		bool	_feature_rescaling = `true`
	)

Is the data sparse Default constructor.

Parameters

[in]	_ml_data	ML Data containing everything!
[in]	feature_rescaling	Feature Rescaling

Note: Default options are to be used when the interface is called from the linear_regression class.

◆ ~linear_regression_opt_interface()

turi::supervised::linear_regression_opt_interface::~linear_regression_opt_interface ( )

Default destructor.

Member Function Documentation

◆ compute_first_order_statistics()

void turi::supervised::linear_regression_opt_interface::compute_first_order_statistics	(	const DenseVector &	point,
		DenseVector &	gradient,
		double &	function_value,
		const size_t	mbStart = `0`,
		const size_t	mbSize = `-1`
	)

virtual

Compute first order statistics at the given point. (Gradient & Function value)

Parameters

[in]	point	Point at which we are computing the stats.
[out]	gradient	Dense gradient
[out]	function_value	Function value
[in]	mbStart	Minibatch start index
[in]	mbSize	Minibatch size (-1 implies all)

Implements turi::optimization::first_order_opt_interface.

◆ compute_function_value()

virtual double turi::optimization::first_order_opt_interface::compute_function_value	(	const DenseVector &	point,
		const size_t	mbStart = `0`,
		const size_t	mbSize = `-1`
	)

virtualinherited

Optimizations for performance reasons.

Compute the function value at a given point.

Parameters

[in]	point	Point at which we are computing the gradient.
[in]	mbStart	Minibatch start index
[in]	mbSize	Minibatch size (-1 implies all)

Returns: Function value at "point"

◆ compute_gradient()

virtual void turi::optimization::first_order_opt_interface::compute_gradient	(	const DenseVector &	point,
		DenseVector &	gradient,
		const size_t	mbStart = `0`,
		const size_t	mbSize = `-1`
	)

virtualinherited

Compute a gradient at the given point.

Parameters

[in]	point	Point at which we are computing the gradient.
[out]	gradient	Dense gradient
[in]	mbStart	Minibatch start index
[in]	mbSize	Minibatch size (-1 implies all)

Warning: See warning on "sparse" compute_gradient to see a potential danger of an infinite loop if you don't imeplement at least one of "dense" or "sparse" versions of compute_gradient.; DO NOT USE this if your model is truly sparse.

◆ compute_hessian()

virtual void turi::optimization::second_order_opt_interface::compute_hessian	(	const DenseVector &	point,
		DenseMatrix &	hessian
	)

virtualinherited

Optimizations for performance reasons.

Compute the hessian at the given point

Parameters

[in]	point	Point at which we are computing the function value.
[out]	hessian	Returns a dense hessian matrix.

Note: Hessians are always dense matrices. Even if they are sparse, it is still really hard to factorize them by make sure the factors stay sparse.

◆ compute_second_order_statistics()

void turi::supervised::linear_regression_opt_interface::compute_second_order_statistics	(	const DenseVector &	point,
		DenseMatrix &	hessian,
		DenseVector &	gradient,
		double &	function_value
	)

virtual

Compute second order statistics at the given point. (Gradient & Function value)

Parameters

[in]	point	Point at which we are computing the stats.
[out]	hessian	Hessian (Dense)
[out]	gradient	Dense gradient
[out]	function_value	Function value

Implements turi::optimization::second_order_opt_interface.

◆ compute_validation_first_order_statistics()

void turi::supervised::linear_regression_opt_interface::compute_validation_first_order_statistics	(	const DenseVector &	point,
		DenseVector &	gradient,
		double &	function_value
	)

Compute first order statistics at the given point with respect to the validation data. (Gradient & Function value)

Parameters

[in]	point	Point at which we are computing the stats.
[out]	gradient	Dense gradient
[out]	function_value	Function value

◆ get_status()

std::vector<std::string> turi::supervised::linear_regression_opt_interface::get_status	(	const DenseVector &	coefs,
		const std::vector< std::string > &	stats
	)

virtual

Get strings needed to print a row of the progress table.

Parameters

[in]	a	vector of model coefficients.
[in]	a	vector of stats to print at the beginning of each row

Reimplemented from turi::optimization::first_order_opt_interface.

◆ get_status_header()

std::vector<std::pair<std::string, size_t> > turi::supervised::linear_regression_opt_interface::get_status_header ( const std::vector< std::string > & stat_names )

virtual

Get strings needed to print the header for the progress table.

Parameters

[in] a vector of strings to print at the beginning of the header.

Reimplemented from turi::optimization::first_order_opt_interface.

◆ init_feature_rescaling()

void turi::supervised::linear_regression_opt_interface::init_feature_rescaling ( )

Set feature scaling

◆ num_examples()

size_t turi::supervised::linear_regression_opt_interface::num_examples ( ) const

virtual

Get the number of examples in the model

Returns: Number of examples

Implements turi::optimization::first_order_opt_interface.

◆ num_validation_examples()

size_t turi::supervised::linear_regression_opt_interface::num_validation_examples ( ) const

Get the number of validation-set examples in the model

Returns: Number of examples

◆ num_variables()

size_t turi::supervised::linear_regression_opt_interface::num_variables ( ) const

virtual

Get the number of variables in the model

Returns: Number of variables

Implements turi::optimization::first_order_opt_interface.

◆ rescale_solution()

void turi::supervised::linear_regression_opt_interface::rescale_solution ( DenseVector & coefs )

Transform the final solution back to the original scale.

Parameters

[in,out] coefs Solution vector

◆ reset()

virtual void turi::optimization::first_order_opt_interface::reset ( int seed )

virtualinherited

Reset the state of the model's "randomness" source.

Parameters

[in] seed Seed that is the source of randomness.

◆ set_threads()

void turi::supervised::linear_regression_opt_interface::set_threads ( size_t _n_threads )

Set the number of threads.

Parameters

[in] _n_threads Set the number of threads.

Member Data Documentation

◆ examples

size_t turi::supervised::linear_regression_opt_interface::examples = 0

protected

Number of examples

Definition at line 46 of file linear_regression_opt_interface.hpp.

◆ feature_rescaling

bool turi::supervised::linear_regression_opt_interface::feature_rescaling = false

protected

<Scale features

Definition at line 52 of file linear_regression_opt_interface.hpp.

◆ features

size_t turi::supervised::linear_regression_opt_interface::features = 0

protected

Number of features

Definition at line 47 of file linear_regression_opt_interface.hpp.

◆ is_dense

bool turi::supervised::linear_regression_opt_interface::is_dense = false

protected

Feature rescaling

Definition at line 53 of file linear_regression_opt_interface.hpp.

◆ scaler

std::shared_ptr<l2_rescaling> turi::supervised::linear_regression_opt_interface::scaler

protected

< Num threads

Definition at line 51 of file linear_regression_opt_interface.hpp.

◆ variables

size_t turi::supervised::linear_regression_opt_interface::variables = 0

protected

Number of variables

Definition at line 48 of file linear_regression_opt_interface.hpp.

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

toolkits/supervised_learning/linear_regression_opt_interface.hpp

Public Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

◆ linear_regression_opt_interface()

◆ ~linear_regression_opt_interface()

Member Function Documentation

◆ compute_first_order_statistics()

◆ compute_function_value()

Optimizations for performance reasons.

◆ compute_gradient()

◆ compute_hessian()

Optimizations for performance reasons.

◆ compute_second_order_statistics()

◆ compute_validation_first_order_statistics()

◆ get_status()

◆ get_status_header()

◆ init_feature_rescaling()

◆ num_examples()

◆ num_validation_examples()

◆ num_variables()

◆ rescale_solution()

◆ reset()

◆ set_threads()

Member Data Documentation

◆ examples

◆ feature_rescaling

◆ features

◆ is_dense

◆ scaler

◆ variables