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

Dmatrix refactor stage 1 (#3301)

Browse Source 
* Use sparse page as singular CSR matrix representation

* Simplify dmatrix methods

* Reduce statefullness of batch iterators

* BREAKING CHANGE: Remove prob_buffer_row parameter. Users are instead recommended to sample their dataset as a preprocessing step before using XGBoost.
This commit is contained in:
Rory Mitchell
2018-06-07 10:25:58 +12:00
committed by GitHub
parent 286dccb8e8
commit a96039141a
47 changed files with 650 additions and 1036 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
include/xgboost/learner.h
View File

@@ -167,7 +167,7 @@ class Learner : public rabit::Serializable {
* \param out_preds output vector to hold the predictions
* \param ntree_limit limit the number of trees used in prediction
*/
inline void Predict(const SparseBatch::Inst &inst,
inline void Predict(const SparsePage::Inst &inst,
bool output_margin,
HostDeviceVector<bst_float> *out_preds,
unsigned ntree_limit = 0) const;
@@ -190,7 +190,7 @@ class Learner : public rabit::Serializable {
};
// implementation of inline functions.
inline void Learner::Predict(const SparseBatch::Inst& inst,
inline void Learner::Predict(const SparsePage::Inst& inst,
bool output_margin,
HostDeviceVector<bst_float>* out_preds,
unsigned ntree_limit) const {