lars.hpp

Go to the documentation of this file.

#ifndef MLPACK_METHODS_LARS_LARS_HPP
#define MLPACK_METHODS_LARS_LARS_HPP

#include <mlpack/prereqs.hpp>

namespace mlpack {
namespace regression {

// beta is the estimator
// yHat is the prediction from the current estimator

class LARS
{
 public:
  LARS(const bool useCholesky = false,
       const double lambda1 = 0.0,
       const double lambda2 = 0.0,
       const double tolerance = 1e-16);

  LARS(const bool useCholesky,
       const arma::mat& gramMatrix,
       const double lambda1 = 0.0,
       const double lambda2 = 0.0,
       const double tolerance = 1e-16);

  LARS(const arma::mat& data,
       const arma::rowvec& responses,
       const bool transposeData = true,
       const bool useCholesky = false,
       const double lambda1 = 0.0,
       const double lambda2 = 0.0,
       const double tolerance = 1e-16);

  LARS(const arma::mat& data,
       const arma::rowvec& responses,
       const bool transposeData,
       const bool useCholesky,
       const arma::mat& gramMatrix,
       const double lambda1 = 0.0,
       const double lambda2 = 0.0,
       const double tolerance = 1e-16);

  LARS(const LARS& other);

  LARS(LARS&& other);

  LARS& operator=(const LARS& other);

  LARS& operator=(LARS&& other);

  double Train(const arma::mat& data,
               const arma::rowvec& responses,
               arma::vec& beta,
               const bool transposeData = true);

  double Train(const arma::mat& data,
               const arma::rowvec& responses,
               const bool transposeData = true);

  void Predict(const arma::mat& points,
               arma::rowvec& predictions,
               const bool rowMajor = false) const;

  const std::vector<size_t>& ActiveSet() const { return activeSet; }

  const std::vector<arma::vec>& BetaPath() const { return betaPath; }

  const arma::vec& Beta() const { return betaPath.back(); }

  const std::vector<double>& LambdaPath() const { return lambdaPath; }

  const arma::mat& MatUtriCholFactor() const { return matUtriCholFactor; }

  template<typename Archive>
  void serialize(Archive& ar, const unsigned int /* version */);

  double ComputeError(const arma::mat& matX,
                      const arma::rowvec& y,
                      const bool rowMajor = false);

 private:
  arma::mat matGramInternal;

  const arma::mat* matGram;

  arma::mat matUtriCholFactor;

  bool useCholesky;

  bool lasso;
  double lambda1;

  bool elasticNet;
  double lambda2;

  double tolerance;

  std::vector<arma::vec> betaPath;

  std::vector<double> lambdaPath;

  std::vector<size_t> activeSet;

  std::vector<bool> isActive;

  // Set of variables that are ignored (if any).

  std::vector<size_t> ignoreSet;

  std::vector<bool> isIgnored;

  void Deactivate(const size_t activeVarInd);

  void Activate(const size_t varInd);

  void Ignore(const size_t varInd);

  // compute "equiangular" direction in output space
  void ComputeYHatDirection(const arma::mat& matX,
                            const arma::vec& betaDirection,
                            arma::vec& yHatDirection);

  // interpolate to compute last solution vector
  void InterpolateBeta();

  void CholeskyInsert(const arma::vec& newX, const arma::mat& X);

  void CholeskyInsert(double sqNormNewX, const arma::vec& newGramCol);

  void GivensRotate(const arma::vec::fixed<2>& x,
                    arma::vec::fixed<2>& rotatedX,
                    arma::mat& G);

  void CholeskyDelete(const size_t colToKill);
};

} // namespace regression
} // namespace mlpack

// Include implementation of serialize().
#include "lars_impl.hpp"

#endif