init commit
This commit is contained in:
tqchen
13
booster/xgboost.cpp
Normal file
13
booster/xgboost.cpp
Normal file
@@ -0,0 +1,13 @@
|
||||
/*!
|
||||
* \file xgboost.cpp
|
||||
* \brief bootser implementations
|
||||
* \author Tianqi Chen: tianqi.tchen@gmail.com
|
||||
*/
|
||||
// implementation of boosters go to here
|
||||
#define _CRT_SECURE_NO_WARNINGS
|
||||
#define _CRT_SECURE_NO_DEPRECATE
|
||||
#include <climits>
|
||||
#include "xgboost.h"
|
||||
#include "../utils/xgboost_utils.h"
|
||||
|
||||
|
||||
11
booster/xgboost.cpp~
Normal file
11
booster/xgboost.cpp~
Normal file
@@ -0,0 +1,11 @@
|
||||
/*!
|
||||
* \file apex_booster.cpp
|
||||
* \brief bootser implementation
|
||||
* \author Tianqi Chen: tqchen@apex.sjtu.edu.cn
|
||||
*/
|
||||
// implementation of boosters go to here
|
||||
#define _CRT_SECURE_NO_WARNINGS
|
||||
#define _CRT_SECURE_NO_DEPRECATE
|
||||
#include <climits>
|
||||
#include "apex_booster.h"
|
||||
#include "../../apex-utils/apex_utils.h"
|
||||
96
booster/xgboost.h
Normal file
96
booster/xgboost.h
Normal file
@@ -0,0 +1,96 @@
|
||||
#ifndef _XGBOOST_H_
|
||||
#define _XGBOOST_H_
|
||||
/*!
|
||||
* \file xgboost.h
|
||||
* \brief the general gradient boosting interface
|
||||
* \author Tianqi Chen: tianqi.tchen@gmail.com
|
||||
*/
|
||||
#include <vector>
|
||||
#include "../utils/xgboost_utils.h"
|
||||
#include "../utils/xgboost_stream.h"
|
||||
#include "xgboost_data.h"
|
||||
|
||||
/*! \brief namespace for xboost package */
|
||||
namespace xgboost{
|
||||
namespace booster{
|
||||
/*! \brief interface of a gradient boosting learner */
|
||||
class IBooster{
|
||||
public:
|
||||
// interface for model setting and loading
|
||||
// calling procedure:
|
||||
// (1) booster->SetParam to setting necessary parameters
|
||||
// (2) if it is first time usage of the model: call booster->
|
||||
// if new model to be trained, trainer->init_trainer
|
||||
// elseif just to load from file, trainer->load_model
|
||||
// trainer->do_boost
|
||||
// trainer->save_model
|
||||
/*!
|
||||
* \brief set parameters from outside
|
||||
* \param name name of the parameter
|
||||
* \param val value of the parameter
|
||||
*/
|
||||
virtual void SetParam( const char *name, const char *val ) = 0;
|
||||
/*!
|
||||
* \brief load model from stream
|
||||
* \param fi input stream
|
||||
*/
|
||||
virtual void LoadModel( utils::IStream &fi ) = 0;
|
||||
/*!
|
||||
* \brief save model to stream
|
||||
* \param fo output stream
|
||||
*/
|
||||
virtual void SaveModel( utils::IStream &fo ) const = 0;
|
||||
/*!
|
||||
* \brief initialize solver before training, called before training
|
||||
* this function is reserved for solver to allocate necessary space and do other preparations
|
||||
*/
|
||||
virtual void InitModel( void ) = 0;
|
||||
public:
|
||||
/*!
|
||||
* \brief do gradient boost training for one step, using the information given
|
||||
* \param grad first order gradient of each instance
|
||||
* \param hess second order gradient of each instance
|
||||
* \param feats features of each instance
|
||||
* \param root_index pre-partitioned root index of each instance,
|
||||
* root_index.size() can be 0 which indicates that no pre-partition involved
|
||||
*/
|
||||
virtual void DoBoost( std::vector<float> &grad,
|
||||
std::vector<float> &hess,
|
||||
const FMatrixS::Image &feats,
|
||||
const std::vector<unsigned> &root_index ) = 0;
|
||||
/*!
|
||||
* \brief predict values for given sparse feature
|
||||
* NOTE: in tree implementation, this is not threadsafe
|
||||
* \param feat vector in sparse format
|
||||
* \param rid root id of current instance, default = 0
|
||||
* \return prediction
|
||||
*/
|
||||
virtual float Predict( const FMatrixS::Line &feat, unsigned rid = 0 ){
|
||||
utils::Error( "not implemented" );
|
||||
return 0.0f;
|
||||
}
|
||||
/*!
|
||||
* \brief predict values for given dense feature
|
||||
* \param feat feature vector in dense format
|
||||
* \param funknown indicator that the feature is missing
|
||||
* \param rid root id of current instance, default = 0
|
||||
* \return prediction
|
||||
*/
|
||||
virtual float Predict( const std::vector<float> &feat,
|
||||
const std::vector<bool> &funknown,
|
||||
unsigned rid = 0 ){
|
||||
utils::Error( "not implemented" );
|
||||
return 0.0f;
|
||||
}
|
||||
/*!
|
||||
* \brief print information
|
||||
* \param fo output stream
|
||||
*/
|
||||
virtual void PrintInfo( FILE *fo ){}
|
||||
public:
|
||||
virtual ~IBooster( void ){}
|
||||
};
|
||||
};
|
||||
};
|
||||
|
||||
#endif
|
||||
93
booster/xgboost.h~
Normal file
93
booster/xgboost.h~
Normal file
@@ -0,0 +1,93 @@
|
||||
#ifndef _XGBOOST_H_
|
||||
#define _XGBOOST_H_
|
||||
/*!
|
||||
* \file xgboost.h
|
||||
* \brief the general gradient boosting interface
|
||||
* \author Tianqi Chen: tianqi.tchen@gmail.com
|
||||
*/
|
||||
|
||||
#include "../utils/xgboost_utils.h"
|
||||
/*! \brief namespace for xboost package */
|
||||
namespace xgboost{
|
||||
namespace booster{
|
||||
/*! \brief interface of a gradient boosting learner */
|
||||
class IBooster{
|
||||
public:
|
||||
// interface for model setting and loading
|
||||
// calling procedure:
|
||||
// (1) booster->SetParam to setting necessary parameters
|
||||
// (2) if it is first time usage of the model: call booster->
|
||||
// if new model to be trained, trainer->init_trainer
|
||||
// elseif just to load from file, trainer->load_model
|
||||
// trainer->do_boost
|
||||
// trainer->save_model
|
||||
/*!
|
||||
* \brief set parameters from outside
|
||||
* \param name name of the parameter
|
||||
* \param val value of the parameter
|
||||
*/
|
||||
virtual void SetParam( const char *name, const char *val ) = 0;
|
||||
/*!
|
||||
* \brief load model from stream
|
||||
* \param fi input stream
|
||||
*/
|
||||
virtual void LoadModel( utils::IStream &fi ) = 0;
|
||||
/*!
|
||||
* \brief save model to stream
|
||||
* \param fo output stream
|
||||
*/
|
||||
virtual void SaveModel( utils::IStream &fo ) const = 0;
|
||||
/*!
|
||||
* \brief initialize solver before training, called before training
|
||||
* this function is reserved for solver to allocate necessary space and do other preparations
|
||||
*/
|
||||
virtual void InitModel( void ) = 0;
|
||||
public:
|
||||
/*!
|
||||
* \brief do gradient boost training for one step, using the information given
|
||||
* \param grad first order gradient of each instance
|
||||
* \param hess second order gradient of each instance
|
||||
* \param feats features of each instance
|
||||
* \param root_index pre-partitioned root index of each instance,
|
||||
* root_index.size() can be 0 which indicates that no pre-partition involved
|
||||
*/
|
||||
virtual void DoBoost( std::vector<float> &grad,
|
||||
std::vector<float> &hess,
|
||||
const FMatrixS::Image &feats,
|
||||
const std::vector<unsigned> &root_index ) = 0;
|
||||
/*!
|
||||
* \brief predict values for given sparse feature
|
||||
* NOTE: in tree implementation, this is not threadsafe
|
||||
* \param feat vector in sparse format
|
||||
* \param rid root id of current instance, default = 0
|
||||
* \return prediction
|
||||
*/
|
||||
virtual float Predict( const FMatrixS::Line &feat, unsigned rid = 0 ){
|
||||
utils::error( "not implemented" );
|
||||
return 0.0f;
|
||||
}
|
||||
/*!
|
||||
* \brief predict values for given dense feature
|
||||
* \param feat feature vector in dense format
|
||||
* \param funknown indicator that the feature is missing
|
||||
* \param rid root id of current instance, default = 0
|
||||
* \return prediction
|
||||
*/
|
||||
virtual float Predict( const std::vector<float> &feat,
|
||||
const std::vector<bool> &funknown,
|
||||
unsigned rid = 0 ){
|
||||
utils::error( "not implemented" );
|
||||
return 0.0f;
|
||||
}
|
||||
/*!
|
||||
* \brief print information
|
||||
* \param fo output stream
|
||||
*/
|
||||
virtual void PrintInfo( FILE *fo ){}
|
||||
public:
|
||||
virtual ~IBooster( void ){}
|
||||
};
|
||||
};
|
||||
};
|
||||
|
||||
#endif
|
||||
118
booster/xgboost_data.h
Normal file
118
booster/xgboost_data.h
Normal file
@@ -0,0 +1,118 @@
|
||||
#ifndef _XGBOOST_DATA_H_
|
||||
#define _XGBOOST_DATA_H_
|
||||
/*!
|
||||
* \file xgboost_data.h
|
||||
* \brief the input data structure for gradient boosting
|
||||
* \author Tianqi Chen: tianqi.tchen@gmail.com
|
||||
*/
|
||||
|
||||
#include <vector>
|
||||
#include "../utils/xgboost_utils.h"
|
||||
|
||||
namespace xgboost{
|
||||
namespace booster{
|
||||
/*! \brief interger type used in boost */
|
||||
typedef int bst_int;
|
||||
/*! \brief unsigned interger type used in boost */
|
||||
typedef unsigned bst_uint;
|
||||
/*! \brief float type used in boost */
|
||||
typedef float bst_float;
|
||||
/*! \brief debug option for booster */
|
||||
const bool bst_debug = false;
|
||||
};
|
||||
};
|
||||
namespace xgboost{
|
||||
namespace booster{
|
||||
/*!
|
||||
* \brief auxlilary feature matrix to store training instance, in sparse CSR format
|
||||
*/
|
||||
class FMatrixS{
|
||||
public:
|
||||
/*! \brief one row of sparse feature matrix */
|
||||
struct Line{
|
||||
/*! \brief array of feature index */
|
||||
const bst_uint *findex;
|
||||
/*! \brief array of feature value */
|
||||
const bst_float *fvalue;
|
||||
/*! \brief size of the data */
|
||||
bst_int len;
|
||||
};
|
||||
/*!
|
||||
* \brief remapped image of sparse matrix,
|
||||
* allows use a subset of sparse matrix, by specifying a rowmap
|
||||
*/
|
||||
struct Image{
|
||||
public:
|
||||
Image( const FMatrixS &smat ):smat(smat), row_map( tmp_rowmap ){
|
||||
}
|
||||
Image( const FMatrixS &smat, const std::vector<unsigned> &row_map )
|
||||
:smat(smat), row_map(row_map){
|
||||
}
|
||||
/*! \brief get sparse part of current row */
|
||||
inline Line operator[]( size_t sidx ) const{
|
||||
if( row_map.size() == 0 ) return smat[ sidx ];
|
||||
else return smat[ row_map[ sidx ] ];
|
||||
}
|
||||
private:
|
||||
// used to set the simple case
|
||||
std::vector<unsigned> tmp_rowmap;
|
||||
const FMatrixS &smat;
|
||||
const std::vector<unsigned> &row_map;
|
||||
};
|
||||
public:
|
||||
// -----Note: unless needed for hacking, these fields should not be accessed directly -----
|
||||
/*! \brief row pointer of CSR sparse storage */
|
||||
std::vector<size_t> row_ptr;
|
||||
/*! \brief index of CSR format */
|
||||
std::vector<bst_uint> findex;
|
||||
/*! \brief value of CSR format */
|
||||
std::vector<bst_float> fvalue;
|
||||
public:
|
||||
/*! \brief constructor */
|
||||
FMatrixS( void ){ this->Clear(); }
|
||||
/*!
|
||||
* \brief get number of rows
|
||||
* \return number of rows
|
||||
*/
|
||||
inline size_t NumRow( void ) const{
|
||||
return row_ptr.size() - 1;
|
||||
}
|
||||
/*! \brief clear the storage */
|
||||
inline void Clear( void ){
|
||||
row_ptr.resize( 0 );
|
||||
findex.resize( 0 );
|
||||
fvalue.resize( 0 );
|
||||
row_ptr.push_back( 0 );
|
||||
}
|
||||
/*!
|
||||
* \brief add a row to the matrix, but only accept features from fstart to fend
|
||||
* \param feat sparse feature
|
||||
* \param fstart start bound of feature
|
||||
* \param fend end bound range of feature
|
||||
* \return the row id addted
|
||||
*/
|
||||
inline size_t AddRow( const Line &feat, unsigned fstart = 0, unsigned fend = UINT_MAX ){
|
||||
utils::Assert( feat.len >= 0, "sparse feature length can not be negative" );
|
||||
unsigned cnt = 0;
|
||||
for( int i = 0; i < feat.len; i ++ ){
|
||||
if( feat.findex[i] < fstart || feat.findex[i] >= fend ) continue;
|
||||
findex.push_back( feat.findex[i] );
|
||||
fvalue.push_back( feat.fvalue[i] );
|
||||
cnt ++;
|
||||
}
|
||||
row_ptr.push_back( row_ptr.back() + cnt );
|
||||
return row_ptr.size() - 2;
|
||||
}
|
||||
/*! \brief get sparse part of current row */
|
||||
inline Line operator[]( size_t sidx ) const{
|
||||
Line sp;
|
||||
utils::Assert( !bst_debug || sidx < this->NumRow(), "row id exceed bound" );
|
||||
sp.len = row_ptr[ sidx + 1 ] - row_ptr[ sidx ];
|
||||
sp.findex = &findex[ row_ptr[ sidx ] ];
|
||||
sp.fvalue = &fvalue[ row_ptr[ sidx ] ];
|
||||
return sp;
|
||||
}
|
||||
};
|
||||
};
|
||||
};
|
||||
#endif
|
||||
118
booster/xgboost_data.h~
Normal file
118
booster/xgboost_data.h~
Normal file
@@ -0,0 +1,118 @@
|
||||
#ifndef _XGBOOST_DATA_H_
|
||||
#define _XGBOOST_DATA_H_
|
||||
/*!
|
||||
* \file xgboost_data.h
|
||||
* \brief the input data structure for gradient boosting
|
||||
* \author Tianqi Chen: tianqi.tchen@gmail.com
|
||||
*/
|
||||
|
||||
#include <vector>
|
||||
#include "../utils/xgboost_utils.h"
|
||||
|
||||
namespace xgboost{
|
||||
namespace booster{
|
||||
/*! \brief interger type used in boost */
|
||||
typedef int bst_int;
|
||||
/*! \brief unsigned interger type used in boost */
|
||||
typedef unsigned bst_uint;
|
||||
/*! \brief float type used in boost */
|
||||
typedef float bst_float;
|
||||
/*! \brief debug option for booster */
|
||||
const bool bst_debug = false;
|
||||
};
|
||||
};
|
||||
namespace xgboost{
|
||||
namespace booster{
|
||||
/*!
|
||||
* \brief auxlilary feature matrix to store training instance, in sparse CSR format
|
||||
*/
|
||||
class FMatrixS{
|
||||
public:
|
||||
/*! \brief one row of sparse feature matrix */
|
||||
struct Line{
|
||||
/*! \brief array of feature index */
|
||||
const bst_uint *findex;
|
||||
/*! \brief array of feature value */
|
||||
const bst_float *fvalue;
|
||||
/*! \brief size of the data */
|
||||
bst_int len;
|
||||
};
|
||||
/*!
|
||||
* \brief remapped image of sparse matrix,
|
||||
* allows use a subset of sparse matrix, by specifying a rowmap
|
||||
*/
|
||||
struct Image{
|
||||
public:
|
||||
Image( const FMatrixS &smat ):smat(smat), row_map( tmp_rowmap ){
|
||||
}
|
||||
Image( const FMatrixS &smat, const std::vector<unsigned> &row_map )
|
||||
:smat(smat), row_map(row_map){
|
||||
}
|
||||
/*! \brief get sparse part of current row */
|
||||
inline Line operator[]( size_t sidx ) const{
|
||||
if( row_map.size() == 0 ) return smat[ sidx ];
|
||||
else return smat[ row_map[ sidx ] ];
|
||||
}
|
||||
private:
|
||||
// used to set the simple case
|
||||
std::vector<unsigned> tmp_rowmap;
|
||||
const FMatrixS &smat;
|
||||
const std::vector<unsigned> &row_map;
|
||||
};
|
||||
public:
|
||||
// -----Note: unless needed for hacking, these fields should not be accessed directly -----
|
||||
/*! \brief row pointer of CSR sparse storage */
|
||||
std::vector<size_t> row_ptr;
|
||||
/*! \brief index of CSR format */
|
||||
std::vector<bst_uint> findex;
|
||||
/*! \brief value of CSR format */
|
||||
std::vector<bst_float> fvalue;
|
||||
public:
|
||||
/*! \brief constructor */
|
||||
FMatrixS( void ){ this->clear(); }
|
||||
/*!
|
||||
* \brief get number of rows
|
||||
* \return number of rows
|
||||
*/
|
||||
inline size_t NumRow( void ) const{
|
||||
return row_ptr.size() - 1;
|
||||
}
|
||||
/*! \brief clear the storage */
|
||||
inline void Clear( void ){
|
||||
row_ptr.resize( 0 );
|
||||
findex.resize( 0 );
|
||||
fvalue.resize( 0 );
|
||||
row_ptr.push_back( 0 );
|
||||
}
|
||||
/*!
|
||||
* \brief add a row to the matrix, but only accept features from fstart to fend
|
||||
* \param feat sparse feature
|
||||
* \param fstart start bound of feature
|
||||
* \param fend end bound range of feature
|
||||
* \return the row id addted
|
||||
*/
|
||||
inline size_t AddRow( const Line &feat, unsigned fstart = 0, unsigned fend = UINT_MAX ){
|
||||
utils::assert( feat.len >= 0, "sparse feature length can not be negative" );
|
||||
unsigned cnt = 0;
|
||||
for( int i = 0; i < feat.len; i ++ ){
|
||||
if( feat.findex[i] < fstart || feat.findex[i] >= fend ) continue;
|
||||
findex.push_back( feat.findex[i] );
|
||||
fvalue.push_back( feat.fvalue[i] );
|
||||
cnt ++;
|
||||
}
|
||||
row_ptr.push_back( row_ptr.back() + cnt );
|
||||
return row_ptr.size() - 2;
|
||||
}
|
||||
/*! \brief get sparse part of current row */
|
||||
inline Line operator[]( size_t sidx ) const{
|
||||
Line sp;
|
||||
utils::assert( !bst_debug || sidx < num_row(), "row id exceed bound" );
|
||||
sp.len = row_ptr[ sidx + 1 ] - row_ptr[ sidx ];
|
||||
sp.findex = &findex[ row_ptr[ sidx ] ];
|
||||
sp.fvalue = &fvalue[ row_ptr[ sidx ] ];
|
||||
return sp;
|
||||
}
|
||||
};
|
||||
};
|
||||
};
|
||||
#endif
|
||||
Reference in New Issue
Block a user