forpy  2
forpy::TreeDesk Struct Reference

Desk for tree training. More...

#include <desk.h>

Public Member Functions

void setup (std::atomic< size_t > *silp, std::atomic< id_t > *np, std::vector< std::pair< id_t, id_t >> *tsp)
 Set up all the internal pointers. More...
 
void reset ()
 Clear the marks and reset all pointers. More...
 

Public Attributes

std::vector< TodoMarkmarks
 
std::atomic< size_t > * stored_in_leafs
 
std::atomic< id_t > * next_id_p
 
std::vector< std::pair< id_t, id_t > > * tree_p
 

Detailed Description

Desk for tree training.

Storage container for threaded tree training (see also Desk Implementations).

Definition at line 19 of file desk.h.

Member Function Documentation

◆ reset()

void forpy::TreeDesk::reset ( )
inline

Clear the marks and reset all pointers.

Definition at line 45 of file desk.h.

◆ setup()

void forpy::TreeDesk::setup ( std::atomic< size_t > *  silp,
std::atomic< id_t > *  np,
std::vector< std::pair< id_t, id_t >> *  tsp 
)
inline

Set up all the internal pointers.

Definition at line 36 of file desk.h.

Member Data Documentation

◆ marks

std::vector<TodoMark> forpy::TreeDesk::marks

Storing marked nodes for DFS. (std::stack has no clear() method, that's why I gave preference to an std::vector.)

Definition at line 22 of file desk.h.

◆ next_id_p

std::atomic<id_t>* forpy::TreeDesk::next_id_p

To determine the ID of the next tree node to use. Usually points to forpy::Tree::next_id (see the doc there for the storage concept).

Definition at line 28 of file desk.h.

◆ stored_in_leafs

std::atomic<size_t>* forpy::TreeDesk::stored_in_leafs

Tracking how many samples have been stored (mainly as sanity check). Usually points to forpy::Tree::stored_in_leafs.

Definition at line 25 of file desk.h.

◆ tree_p

std::vector<std::pair<id_t, id_t> >* forpy::TreeDesk::tree_p

A vector representation of the tree. Usually points to forpy::Tree::tree.

Definition at line 31 of file desk.h.


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