slope::SlopeFit Class Reference

A class representing the results of SLOPE (Sorted L1 Penalized Estimation) fitting. More...

#include <slope_fit.h>

Public Member Functions
	SlopeFit (const Eigen::VectorXd &intercepts, const Eigen::SparseMatrix< double > &coefs, const Clusters &clusters, const double alpha, const Eigen::ArrayXd &lambda, const double deviance, const double null_deviance, const std::vector< double > &primals, const std::vector< double > &duals, const std::vector< double > &time, const int passes, const std::string &centering_type, const std::string &scaling_type)
	Construct a new Slope Fit object.
const Eigen::VectorXd &	getIntercepts () const
	Gets the intercept terms for this SLOPE fit.
const Eigen::SparseMatrix< double > &	getCoefs () const
	Gets the sparse coefficient matrix for this fit.
const Clusters &	getClusters () const
	Gets the clusters.
const Eigen::ArrayXd &	getLambda () const
	Gets the lambda (regularization) parameter used.
double	getAlpha () const
	Gets the alpha (mixing) parameter used.
double	getDeviance () const
	Gets the model deviance.
const std::string &	getLossType () const
	Gets the model's loss type.
double	getNullDeviance () const
	Gets the null model deviance.
const std::vector< double > &	getPrimals () const
	Gets the sequence of primal objective values during optimization.
const std::vector< double > &	getDuals () const
	Gets the sequence of dual objective values during optimization.
const std::vector< double > &	getTime () const
	Gets the sequence of computation times during optimization.
int	getPasses () const
	Gets the total number of optimization iterations.
double	getDevianceRatio () const
	Calculate the deviance ratio (1 - deviance/null_deviance)
std::vector< double >	getGaps () const
	Calculate the duality gaps during optimization.
template<typename T >
Eigen::MatrixXd	predict (T &x, const std::string &type="response") const
	Predict the response for a given input matrix.

Detailed Description

A class representing the results of SLOPE (Sorted L1 Penalized Estimation) fitting.

This class stores the results of a SLOPE regression, including coefficients, intercepts, clusters (if required), regularization parameters, and optimization metrics.

Definition at line 26 of file slope_fit.h.

Constructor & Destructor Documentation

SlopeFit()

slope::SlopeFit::SlopeFit ( const Eigen::VectorXd &  intercepts, const Eigen::SparseMatrix< double > &  coefs, const Clusters &  clusters, const double  alpha, const Eigen::ArrayXd &  lambda, const double  deviance, const double  null_deviance, const std::vector< double > &  primals, const std::vector< double > &  duals, const std::vector< double > &  time, const int  passes, const std::string &  centering_type, const std::string &  scaling_type  )

inline

Construct a new Slope Fit object.

Parameters

intercepts	Vector of intercept terms
coefs	Matrix of fitted coefficients
clusters	Clusters of coefficients
alpha	Mixing parameter between L1 and SLOPE norms
lambda	Sequence of decreasing weights
deviance	Final model deviance
null_deviance	Null model deviance
primals	History of primal objectives
duals	History of dual objectives
time	Vector of optimization timestamps
passes	Number of optimization passes
centering_type	Type of centering for the design matrix
scaling_type	Type of scaling for the design matrix

Definition at line 67 of file slope_fit.h.

Member Function Documentation

getAlpha()

double slope::SlopeFit::getAlpha ( ) const

inline

Gets the alpha (mixing) parameter used.

Definition at line 119 of file slope_fit.h.

getClusters()

const Clusters & slope::SlopeFit::getClusters ( ) const

inline

Gets the clusters.

Definition at line 109 of file slope_fit.h.

getCoefs()

const Eigen::SparseMatrix< double > & slope::SlopeFit::getCoefs ( ) const

inline

Gets the sparse coefficient matrix for this fit.

Definition at line 104 of file slope_fit.h.

getDeviance()

double slope::SlopeFit::getDeviance ( ) const

inline

Gets the model deviance.

Definition at line 124 of file slope_fit.h.

getDevianceRatio()

double slope::SlopeFit::getDevianceRatio ( ) const

inline

Calculate the deviance ratio (1 - deviance/null_deviance)

Returns

double The deviance ratio, a measure of model fit (higher is better)

Definition at line 162 of file slope_fit.h.

getDuals()

const std::vector< double > & slope::SlopeFit::getDuals ( ) const

inline

Gets the sequence of dual objective values during optimization.

Definition at line 144 of file slope_fit.h.

getGaps()

std::vector< double > slope::SlopeFit::getGaps ( ) const

inline

Calculate the duality gaps during optimization.

Returns

std::vector<double> Vector of duality gaps (primal - dual objectives)

Definition at line 170 of file slope_fit.h.

getIntercepts()

const Eigen::VectorXd & slope::SlopeFit::getIntercepts ( ) const

inline

Gets the intercept terms for this SLOPE fit.

Definition at line 99 of file slope_fit.h.

getLambda()

const Eigen::ArrayXd & slope::SlopeFit::getLambda ( ) const

inline

Gets the lambda (regularization) parameter used.

Definition at line 114 of file slope_fit.h.

getLossType()

const std::string & slope::SlopeFit::getLossType ( ) const

inline

Gets the model's loss type.

Definition at line 129 of file slope_fit.h.

getNullDeviance()

double slope::SlopeFit::getNullDeviance ( ) const

inline

Gets the null model deviance.

Definition at line 134 of file slope_fit.h.

getPasses()

int slope::SlopeFit::getPasses ( ) const

inline

Gets the total number of optimization iterations.

Definition at line 154 of file slope_fit.h.

getPrimals()

const std::vector< double > & slope::SlopeFit::getPrimals ( ) const

inline

Gets the sequence of primal objective values during optimization.

Definition at line 139 of file slope_fit.h.

getTime()

const std::vector< double > & slope::SlopeFit::getTime ( ) const

inline

Gets the sequence of computation times during optimization.

Definition at line 149 of file slope_fit.h.

predict()

template<typename T >

Eigen::MatrixXd slope::SlopeFit::predict ( T &  x, const std::string &  type = "response"  ) const

inline

Predict the response for a given input matrix.

Template Parameters

T	Type of input matrix (dense or sparse)

Parameters

x	Input matrix of features
type	Type of prediction to return ("response" or "linear")

Returns

Matrix of predicted responses

See also

Loss

Definition at line 188 of file slope_fit.h.

---

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

• src/slope/slope_fit.h