Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Change Dataset::CopySubrow from group-wise to column-wise #3720

Merged
merged 1 commit into from
Jan 25, 2021
Merged

Change Dataset::CopySubrow from group-wise to column-wise #3720

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions include/LightGBM/feature_group.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -232,6 +232,14 @@ class FeatureGroup {
}
}

inline void CopySubrowByCol(const FeatureGroup* full_feature, const data_size_t* used_indices, data_size_t num_used_indices, int fidx) {
if (!is_multi_val_) {
bin_data_->CopySubrow(full_feature->bin_data_.get(), used_indices, num_used_indices);
} else {
multi_bin_data_[fidx]->CopySubrow(full_feature->multi_bin_data_[fidx].get(), used_indices, num_used_indices);
}
}

void AddFeaturesFrom(const FeatureGroup* other, int group_id) {
CHECK(is_multi_val_);
CHECK(other->is_multi_val_);
Expand Down
29 changes: 25 additions & 4 deletions src/io/dataset.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -790,15 +790,36 @@ void Dataset::CopySubrow(const Dataset* fullset,
const data_size_t* used_indices,
data_size_t num_used_indices, bool need_meta_data) {
CHECK_EQ(num_used_indices, num_data_);
OMP_INIT_EX();
#pragma omp parallel for schedule(static)

std::vector<int> group_ids, subfeature_ids;
group_ids.reserve(num_features_);
subfeature_ids.reserve(num_features_);
for (int group = 0; group < num_groups_; ++group) {
if (fullset->IsMultiGroup(group)) {
for (int sub_feature = 0; sub_feature <
fullset->feature_groups_[group]->num_feature_; ++sub_feature) {
group_ids.emplace_back(group);
subfeature_ids.emplace_back(sub_feature);
}
} else {
group_ids.emplace_back(group);
subfeature_ids.emplace_back(-1);
}
}
int num_copy_tasks = static_cast<int>(group_ids.size());

OMP_INIT_EX();
#pragma omp parallel for schedule(dynamic)
for (int task_id = 0; task_id < num_copy_tasks; ++task_id) {
OMP_LOOP_EX_BEGIN();
feature_groups_[group]->CopySubrow(fullset->feature_groups_[group].get(),
used_indices, num_used_indices);
int group = group_ids[task_id];
int subfeature = subfeature_ids[task_id];
feature_groups_[group]->CopySubrowByCol(fullset->feature_groups_[group].get(),
used_indices, num_used_indices, subfeature);
OMP_LOOP_EX_END();
}
OMP_THROW_EX();

if (need_meta_data) {
metadata_.Init(fullset->metadata_, used_indices, num_used_indices);
}
Expand Down