Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge branch 'dev' of https://github.com/tqchen/xgboost into dev

Browse Source 
Conflicts:
	regrank/xgboost_regrank_obj.hpp
This commit is contained in:
kalenhaha
2014-05-07 22:15:59 +08:00
parent 8eae8d956d 833cf29867
commit 8b3fc78999
17 changed files with 8756 additions and 109 deletions
Download Patch File Download Diff File Expand all files Collapse all files

65
regrank/xgboost_regrank.h
View File

@@ -86,6 +86,7 @@ namespace xgboost{
if (!strcmp(name, "silent")) silent = atoi(val);
if (!strcmp(name, "eval_metric")) evaluator_.AddEval(val);
if (!strcmp(name, "objective") ) name_obj_ = val;
if (!strcmp(name, "num_class") ) base_gbm.SetParam("num_booster_group", val );
mparam.SetParam(name, val);
base_gbm.SetParam(name, val);
cfg_.push_back( std::make_pair( std::string(name), std::string(val) ) );
@@ -95,7 +96,13 @@ namespace xgboost{
* this function is reserved for solver to allocate necessary space and do other preparation
*/
inline void InitTrainer(void){
base_gbm.InitTrainer();
if( mparam.num_class != 0 ){
if( name_obj_ != "softmax" ){
name_obj_ = "softmax";
printf("auto select objective=softmax to support multi-class classification\n" );
}
}
base_gbm.InitTrainer();
obj_ = CreateObjFunction( name_obj_.c_str() );
for( size_t i = 0; i < cfg_.size(); ++ i ){
obj_->SetParam( cfg_[i].first.c_str(), cfg_[i].second.c_str() );
@@ -166,9 +173,18 @@ namespace xgboost{
inline void UpdateOneIter(const DMatrix &train){
this->PredictRaw(preds_, train);
obj_->GetGradient(preds_, train.info, base_gbm.NumBoosters(), grad_, hess_);
// do boost
std::vector<unsigned> root_index;
base_gbm.DoBoost(grad_, hess_, train.data, root_index);
if( grad_.size() == train.Size() ){
base_gbm.DoBoost(grad_, hess_, train.data, train.info.root_index);
}else{
int ngroup = base_gbm.NumBoosterGroup();
utils::Assert( grad_.size() == train.Size() * (size_t)ngroup, "BUG: UpdateOneIter: mclass" );
std::vector<float> tgrad( train.Size() ), thess( train.Size() );
for( int g = 0; g < ngroup; ++ g ){
memcpy( &tgrad[0], &grad_[g*tgrad.size()], sizeof(float)*tgrad.size() );
memcpy( &thess[0], &hess_[g*tgrad.size()], sizeof(float)*tgrad.size() );
base_gbm.DoBoost(tgrad, thess, train.data, train.info.root_index, g );
}
}
}
/*!
* \brief evaluate the model for specific iteration
@@ -190,9 +206,14 @@ namespace xgboost{
fprintf(fo, "\n");
fflush(fo);
}
/*! \brief get prediction, without buffering */
inline void Predict(std::vector<float> &preds, const DMatrix &data){
this->PredictRaw(preds,data);
/*!
* \brief get prediction
* \param storage to store prediction
* \param data input data
* \param bst_group booster group we are in
*/
inline void Predict(std::vector<float> &preds, const DMatrix &data, int bst_group = -1){
this->PredictRaw( preds, data, bst_group );
obj_->PredTransform( preds );
}
public:
@@ -241,24 +262,32 @@ namespace xgboost{
base_gbm.InteractRePredict(data.data, j, buffer_offset + j);
}
}
private:
/*! \brief get un-transformed prediction*/
inline void PredictRaw(std::vector<float> &preds, const DMatrix &data){
this->PredictBuffer(preds, data, this->FindBufferOffset(data) );
inline void PredictRaw(std::vector<float> &preds, const DMatrix &data, int bst_group = -1 ){
int buffer_offset = this->FindBufferOffset(data);
if( bst_group < 0 ){
int ngroup = base_gbm.NumBoosterGroup();
preds.resize( data.Size() * ngroup );
for( int g = 0; g < ngroup; ++ g ){
this->PredictBuffer(&preds[ data.Size() * g ], data, buffer_offset, g );
}
}else{
preds.resize( data.Size() );
this->PredictBuffer(&preds[0], data, buffer_offset, bst_group );
}
}
/*! \brief get the un-transformed predictions, given data */
inline void PredictBuffer(std::vector<float> &preds, const DMatrix &data, int buffer_offset){
preds.resize(data.Size());
inline void PredictBuffer(float *preds, const DMatrix &data, int buffer_offset, int bst_group ){
const unsigned ndata = static_cast<unsigned>(data.Size());
if( buffer_offset >= 0 ){
#pragma omp parallel for schedule( static )
for (unsigned j = 0; j < ndata; ++j){
preds[j] = mparam.base_score + base_gbm.Predict(data.data, j, buffer_offset + j);
preds[j] = mparam.base_score + base_gbm.Predict(data.data, j, buffer_offset + j, data.info.GetRoot(j), bst_group );
}
}else
#pragma omp parallel for schedule( static )
for (unsigned j = 0; j < ndata; ++j){
preds[j] = mparam.base_score + base_gbm.Predict(data.data, j, -1);
preds[j] = mparam.base_score + base_gbm.Predict(data.data, j, -1, data.info.GetRoot(j), bst_group );
}{
}
}
@@ -270,14 +299,17 @@ namespace xgboost{
/* \brief type of loss function */
int loss_type;
/* \brief number of features */
int num_feature;
int num_feature;
/* \brief number of class, if it is multi-class classification */
int num_class;
/*! \brief reserved field */
int reserved[16];
int reserved[15];
/*! \brief constructor */
ModelParam(void){
base_score = 0.5f;
loss_type = 0;
num_feature = 0;
num_class = 0;
memset(reserved, 0, sizeof(reserved));
}
/*!
@@ -288,6 +320,7 @@ namespace xgboost{
inline void SetParam(const char *name, const char *val){
if (!strcmp("base_score", name)) base_score = (float)atof(val);
if (!strcmp("loss_type", name)) loss_type = atoi(val);
if (!strcmp("num_class", name)) num_class = atoi(val);
if (!strcmp("bst:num_feature", name)) num_feature = atoi(val);
}
/*!

23
regrank/xgboost_regrank_data.h
View File

@@ -35,11 +35,17 @@ namespace xgboost{
std::vector<unsigned> group_ptr;
/*! \brief weights of each instance, optional */
std::vector<float> weights;
/*! \brief specified root index of each instance, can be used for multi task setting*/
std::vector<unsigned> root_index;
/*! \brief get weight of each instances */
inline float GetWeight( size_t i ) const{
if( weights.size() != 0 ) return weights[i];
if( weights.size() != 0 ) return weights[i];
else return 1.0f;
}
inline float GetRoot( size_t i ) const{
if( root_index.size() != 0 ) return root_index[i];
else return 0;
}
};
public:
/*! \brief feature data content */
@@ -112,7 +118,10 @@ namespace xgboost{
unsigned ngptr;
if( fs.Read(&ngptr, sizeof(unsigned) ) != 0 ){
info.group_ptr.resize( ngptr );
utils::Assert( fs.Read(&info.group_ptr[0], sizeof(unsigned) * ngptr) != 0, "Load group file");
if( ngptr != 0 ){
utils::Assert( fs.Read(&info.group_ptr[0], sizeof(unsigned) * ngptr) != 0, "Load group file");
utils::Assert( info.group_ptr.back() == data.NumRow(), "number of group must match number of record" );
}
}
}
fs.Close();
@@ -121,7 +130,7 @@ namespace xgboost{
printf("%ux%u matrix with %lu entries is loaded from %s\n",
(unsigned)data.NumRow(), (unsigned)data.NumCol(), (unsigned long)data.NumEntry(), fname);
if( info.group_ptr.size() != 0 ){
printf("data contains %u groups\n", (unsigned)info.group_ptr.size() );
printf("data contains %u groups\n", (unsigned)info.group_ptr.size()-1 );
}
}
this->TryLoadWeight(fname, silent);
@@ -141,16 +150,18 @@ namespace xgboost{
utils::Assert( info.labels.size() == data.NumRow(), "label size is not consistent with feature matrix size" );
fs.Write(&info.labels[0], sizeof(float) * data.NumRow());
{// write out group ptr
unsigned ngptr = static_cast<unsigned>( info.group_ptr.size() );
unsigned ngptr = static_cast<unsigned>( info.group_ptr.size() );
fs.Write(&ngptr, sizeof(unsigned) );
fs.Write(&info.group_ptr[0], sizeof(unsigned) * ngptr);
if( ngptr != 0 ){
fs.Write(&info.group_ptr[0], sizeof(unsigned) * ngptr);
}
}
fs.Close();
if (!silent){
printf("%ux%u matrix with %lu entries is saved to %s\n",
(unsigned)data.NumRow(), (unsigned)data.NumCol(), (unsigned long)data.NumEntry(), fname);
if( info.group_ptr.size() != 0 ){
printf("data contains %u groups\n", (unsigned)info.group_ptr.size() );
printf("data contains %u groups\n", (unsigned)info.group_ptr.size()-1 );
}
}
}

17
regrank/xgboost_regrank_eval.h
View File

@@ -13,6 +13,7 @@
#include "../utils/xgboost_omp.h"
#include "../utils/xgboost_random.h"
#include "xgboost_regrank_data.h"
#include "xgboost_regrank_utils.h"
namespace xgboost{
namespace regrank{
@@ -31,17 +32,11 @@ namespace xgboost{
virtual ~IEvaluator(void){}
};
inline static bool CmpFirst(const std::pair<float, unsigned> &a, const std::pair<float, unsigned> &b){
return a.first > b.first;
}
inline static bool CmpSecond(const std::pair<float, unsigned> &a, const std::pair<float, unsigned> &b){
return a.second > b.second;
}
/*! \brief RMSE */
struct EvalRMSE : public IEvaluator{
virtual float Eval(const std::vector<float> &preds,
const DMatrix::Info &info) const {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
const unsigned ndata = static_cast<unsigned>(preds.size());
float sum = 0.0, wsum = 0.0;
#pragma omp parallel for reduction(+:sum,wsum) schedule( static )
@@ -62,6 +57,7 @@ namespace xgboost{
struct EvalLogLoss : public IEvaluator{
virtual float Eval(const std::vector<float> &preds,
const DMatrix::Info &info) const {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
const unsigned ndata = static_cast<unsigned>(preds.size());
float sum = 0.0f, wsum = 0.0f;
#pragma omp parallel for reduction(+:sum,wsum) schedule( static )
@@ -106,7 +102,8 @@ namespace xgboost{
/*! \brief Area under curve, for both classification and rank */
struct EvalAuc : public IEvaluator{
virtual float Eval(const std::vector<float> &preds,
const DMatrix::Info &info) const {
const DMatrix::Info &info) const {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
std::vector<unsigned> tgptr(2, 0); tgptr[1] = preds.size();
const std::vector<unsigned> &gptr = info.group_ptr.size() == 0 ? tgptr : info.group_ptr;
utils::Assert(gptr.back() == preds.size(), "EvalAuc: group structure must match number of prediction");
@@ -159,8 +156,10 @@ namespace xgboost{
public:
virtual float Eval(const std::vector<float> &preds,
const DMatrix::Info &info) const {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
const std::vector<unsigned> &gptr = info.group_ptr;
utils::Assert(gptr.size() != 0 && gptr.back() == preds.size(), "EvalAuc: group structure must match number of prediction");
utils::Assert(gptr.size() != 0, "must specify group when constructing rank file");
utils::Assert( gptr.back() == preds.size(), "EvalRanklist: group structure must match number of prediction");
const unsigned ngroup = static_cast<unsigned>(gptr.size() - 1);
double sum_metric = 0.0f;

5
regrank/xgboost_regrank_obj.h
View File

@@ -106,8 +106,9 @@ namespace xgboost{
namespace regrank{
IObjFunction* CreateObjFunction( const char *name ){
if( !strcmp("reg", name ) ) return new RegressionObj();
if( !strcmp("rank", name ) ) return new PairwiseRankObj();
if( !strcmp("softmax", name ) ) return new SoftmaxObj();
if( !strcmp("rank:pairwise", name ) ) return new PairwiseRankObj();
if( !strcmp("rank:softmax", name ) ) return new SoftmaxRankObj();
if( !strcmp("softmax", name ) ) return new SoftmaxMultiClassObj();
utils::Error("unknown objective function type");
return NULL;
}

96
regrank/xgboost_regrank_obj.hpp
View File

@@ -1,7 +1,7 @@
#ifndef XGBOOST_REGRANK_OBJ_HPP
#define XGBOOST_REGRANK_OBJ_HPP
/*!
* \file xgboost_regrank_obj.h
* \file xgboost_regrank_obj.hpp
* \brief implementation of objective functions
* \author Tianqi Chen, Kailong Chen
*/
@@ -9,6 +9,8 @@
#include <vector>
#include <functional>
#include "xgboost_regrank_sample.h"
#include "xgboost_regrank_utils.h"
namespace xgboost{
namespace regrank{
class RegressionObj : public IObjFunction{
@@ -25,6 +27,7 @@ namespace xgboost{
int iter,
std::vector<float> &grad,
std::vector<float> &hess ) {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
grad.resize(preds.size()); hess.resize(preds.size());
const unsigned ndata = static_cast<unsigned>(preds.size());
@@ -53,11 +56,11 @@ namespace xgboost{
namespace regrank{
// simple softmax rak
class SoftmaxObj : public IObjFunction{
class SoftmaxRankObj : public IObjFunction{
public:
SoftmaxObj(void){
SoftmaxRankObj(void){
}
virtual ~SoftmaxObj(){}
virtual ~SoftmaxRankObj(){}
virtual void SetParam(const char *name, const char *val){
}
virtual void GetGradient(const std::vector<float>& preds,
@@ -65,6 +68,7 @@ namespace xgboost{
int iter,
std::vector<float> &grad,
std::vector<float> &hess ) {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
grad.resize(preds.size()); hess.resize(preds.size());
const std::vector<unsigned> &gptr = info.group_ptr;
utils::Assert( gptr.size() != 0 && gptr.back() == preds.size(), "rank loss must have group file" );
@@ -97,23 +101,76 @@ namespace xgboost{
}
virtual const char* DefaultEvalMetric(void) {
return "pre@1";
}
private:
inline static void Softmax( std::vector<float>& rec ){
float wmax = rec[0];
for( size_t i = 1; i < rec.size(); ++ i ){
wmax = std::max( rec[i], wmax );
}
double wsum = 0.0f;
for( size_t i = 0; i < rec.size(); ++ i ){
rec[i] = expf(rec[i]-wmax);
wsum += rec[i];
}
for( size_t i = 0; i < rec.size(); ++ i ){
rec[i] /= wsum;
}
}
};
// simple softmax multi-class classification
class SoftmaxMultiClassObj : public IObjFunction{
public:
SoftmaxMultiClassObj(void){
nclass = 0;
}
virtual ~SoftmaxMultiClassObj(){}
virtual void SetParam(const char *name, const char *val){
if( !strcmp( "num_class", name ) ) nclass = atoi(val);
}
virtual void GetGradient(const std::vector<float>& preds,
const DMatrix::Info &info,
int iter,
std::vector<float> &grad,
std::vector<float> &hess ) {
utils::Assert( nclass != 0, "must set num_class to use softmax" );
utils::Assert( preds.size() == (size_t)nclass * info.labels.size(), "SoftmaxMultiClassObj: label size and pred size does not match" );
grad.resize(preds.size()); hess.resize(preds.size());
const unsigned ndata = static_cast<unsigned>(info.labels.size());
#pragma omp parallel
{
std::vector<float> rec(nclass);
#pragma for schedule(static)
for (unsigned j = 0; j < ndata; ++j){
for( int k = 0; k < nclass; ++ k ){
rec[k] = preds[j + k * ndata];
}
Softmax( rec );
int label = static_cast<int>(info.labels[j]);
utils::Assert( label < nclass, "SoftmaxMultiClassObj: label exceed num_class" );
for( int k = 0; k < nclass; ++ k ){
float p = rec[ k ];
if( label == k ){
grad[j+k*ndata] = p - 1.0f;
}else{
grad[j+k*ndata] = p;
}
hess[j+k*ndata] = 2.0f * p * ( 1.0f - p );
}
}
}
}
virtual void PredTransform(std::vector<float> &preds){
utils::Assert( nclass != 0, "must set num_class to use softmax" );
utils::Assert( preds.size() % nclass == 0, "SoftmaxMultiClassObj: label size and pred size does not match" );
const unsigned ndata = static_cast<unsigned>(preds.size()/nclass);
#pragma omp parallel
{
std::vector<float> rec(nclass);
#pragma for schedule(static)
for (unsigned j = 0; j < ndata; ++j){
for( int k = 0; k < nclass; ++ k ){
rec[k] = preds[j + k * ndata];
}
Softmax( rec );
preds[j] = FindMaxIndex( rec );
}
}
preds.resize( ndata );
}
virtual const char* DefaultEvalMetric(void) {
return "error";
}
private:
int nclass;
};
};
namespace regrank{
@@ -134,6 +191,7 @@ namespace xgboost{
int iter,
std::vector<float> &grad,
std::vector<float> &hess ) {
utils::Assert( preds.size() == info.labels.size(), "label size predict size not match" );
grad.resize(preds.size()); hess.resize(preds.size());
const std::vector<unsigned> &gptr = info.group_ptr;
utils::Assert( gptr.size() != 0 && gptr.back() == preds.size(), "rank loss must have group file" );

43
regrank/xgboost_regrank_utils.h Normal file
View File

@@ -0,0 +1,43 @@
#ifndef XGBOOST_REGRANK_UTILS_H
#define XGBOOST_REGRANK_UTILS_H
/*!
* \file xgboost_regrank_utils.h
* \brief useful helper functions
* \author Tianqi Chen, Kailong Chen
*/
namespace xgboost{
namespace regrank{
// simple helper function to do softmax
inline static void Softmax( std::vector<float>& rec ){
float wmax = rec[0];
for( size_t i = 1; i < rec.size(); ++ i ){
wmax = std::max( rec[i], wmax );
}
double wsum = 0.0f;
for( size_t i = 0; i < rec.size(); ++ i ){
rec[i] = expf(rec[i]-wmax);
wsum += rec[i];
}
for( size_t i = 0; i < rec.size(); ++ i ){
rec[i] /= wsum;
}
}
// simple helper function to do softmax
inline static int FindMaxIndex( std::vector<float>& rec ){
size_t mxid = 0;
for( size_t i = 1; i < rec.size(); ++ i ){
if( rec[i] > rec[mxid] ) mxid = i;
}
return (int)mxid;
}
inline static bool CmpFirst(const std::pair<float, unsigned> &a, const std::pair<float, unsigned> &b){
return a.first > b.first;
}
inline static bool CmpSecond(const std::pair<float, unsigned> &a, const std::pair<float, unsigned> &b){
return a.second > b.second;
}
};
};
#endif