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

Support cpu quantile sketch with column-wise data split (#8742)

Browse Source
This commit is contained in:
Rong Ou
2023-02-04 22:26:24 -08:00
committed by GitHub
parent c1786849e3
commit 66191e9926
15 changed files with 250 additions and 118 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
src/common/quantile.h
View File

@@ -802,6 +802,7 @@ class SketchContainerImpl {
std::vector<bst_row_t> columns_size_;
int32_t max_bins_;
bool use_group_ind_{false};
bool col_split_;
int32_t n_threads_;
bool has_categorical_{false};
Monitor monitor_;
@@ -814,7 +815,7 @@ class SketchContainerImpl {
* \param use_group whether is assigned to group to data instance.
*/
SketchContainerImpl(std::vector<bst_row_t> columns_size, int32_t max_bins,
common::Span<FeatureType const> feature_types, bool use_group,
common::Span<FeatureType const> feature_types, bool use_group, bool col_split,
int32_t n_threads);
static bool UseGroup(MetaInfo const &info) {
@@ -896,6 +897,10 @@ class SketchContainerImpl {
void PushRowPage(SparsePage const &page, MetaInfo const &info, Span<float const> hessian = {});
void MakeCuts(HistogramCuts* cuts);
private:
// Merge all categories from other workers.
void AllreduceCategories();
};
class HostSketchContainer : public SketchContainerImpl<WQuantileSketch<float, float>> {
@@ -904,7 +909,8 @@ class HostSketchContainer : public SketchContainerImpl<WQuantileSketch<float, fl
public:
HostSketchContainer(int32_t max_bins, common::Span<FeatureType const> ft,
std::vector<size_t> columns_size, bool use_group, int32_t n_threads);
std::vector<size_t> columns_size, bool use_group, bool col_split,
int32_t n_threads);
template <typename Batch>
void PushAdapterBatch(Batch const &batch, size_t base_rowid, MetaInfo const &info, float missing);
@@ -1000,9 +1006,9 @@ class SortedSketchContainer : public SketchContainerImpl<WXQuantileSketch<float,
public:
explicit SortedSketchContainer(int32_t max_bins, common::Span<FeatureType const> ft,
std::vector<size_t> columns_size, bool use_group,
std::vector<size_t> columns_size, bool use_group, bool col_split,
int32_t n_threads)
: SketchContainerImpl{columns_size, max_bins, ft, use_group, n_threads} {
: SketchContainerImpl{columns_size, max_bins, ft, use_group, col_split, n_threads} {
monitor_.Init(__func__);
sketches_.resize(columns_size.size());
size_t i = 0;