TreeBooster.h

Go to the documentation of this file.

#ifndef TREE_BOOSTER_H
#define TREE_BOOSTER_H
 
#include "CNum/DataStructs/DataStructs.h"
#include "CNum/Data/Data.h"
#include "CNum/Model/Tree/TreeDefs.h"
 
#include <atomic>
#include <vector>
#include <variant>
#include <future>
#include <cstring>
#include <utility>
 
namespace CNum::Model::Tree {
  class TreeBooster {
  protected:
    TreeBoosterNode *_root;
    int _max_depth;
    int _min_samples;
    double _reg_lambda;
    double _gamma;
    double _weight_decay;
    arena_t *_arena;
    
  private:
    double predict_sample(TreeBoosterNode *node, std::span<double> &sample);
 
    virtual void fit_node_greedy(const CNum::DataStructs::Matrix<double> &X,
                 double *g,
                 double *h,
                 TreeBoosterNode *node,
                 int depth = 0) = 0;
    
    virtual void fit_node_hist(const CNum::DataStructs::Matrix<int> &X,
                   std::shared_ptr<CNum::Data::Shelf[]> shelves,
                   double *g,
                   double *h,
                   DataPartition &partition,
                   const arena_view_t &parent_hist_view,
                   TreeBoosterNode *node,
                   int depth = 0) = 0;
 
    virtual void fit_prep(const CNum::DataStructs::Matrix<int> &X,
              std::shared_ptr<CNum::Data::Shelf[]> shelves,
              double *g,
              double *h,
              DataPartition &partition) = 0;
 
    virtual void fit_prep(const CNum::DataStructs::Matrix<double> &X,
              std::shared_ptr<CNum::Data::Shelf[]> shelves,
              double *g,
              double *h,
              DataPartition &partition) = 0;
    void destruct(TreeBoosterNode *node);

    void copy_hyperparams(const TreeBooster &other) noexcept;
    
    void move(TreeBooster &&other) noexcept;

    TreeBoosterNode *copy_tree(TreeBoosterNode *node) noexcept;

    void copy(const TreeBooster &other) noexcept;
  
  public:
    TreeBooster(arena_t *a = nullptr,
        int md = 5,
        int ms = 3,
        double weight_decay = 0.0,
        double reg_lambda = 1.0,
        double gamma = 0.0);
 

    TreeBooster(const TreeBooster &other) noexcept;

    TreeBooster &operator=(const TreeBooster &other) noexcept;

    TreeBooster(TreeBooster &&other) noexcept;

    TreeBooster &operator=(TreeBooster &&other) noexcept;

    virtual ~TreeBooster();

    void set_root(TreeBoosterNode *root);
 
    virtual void fit(DataMatrix &X,
             std::shared_ptr<CNum::Data::Shelf[]> shelves,
             double *g,
             double *h,
             DataPartition &partition) = 0;

    CNum::DataStructs::Matrix<double> predict(CNum::DataStructs::Matrix<double> &data);

    static size_t partition_data(const CNum::DataStructs::Matrix<int> &X,
                 double *g,
                 double *h,
                 size_t feat,
                 uint8_t bin,
                 const DataPartition &partition);

    static void histogram_subtraction(const arena_view_t &parent_hist_view,
                      arena_view_t &small_hist_view,
                      arena_view_t &large_hist_view);

    arena_view_t init_hist_view(size_t n_data_cols);

    std::string to_json();
  };
};
  
#endif