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

[BLOCKING] Handle empty rows in data iterators correctly (#5929)

Browse Source 
* [jvm-packages] Handle empty rows in data iterators correctly

* Fix clang-tidy error

* last empty row

* Add comments [skip ci]

Co-authored-by: Nan Zhu <nanzhu@uber.com>
This commit is contained in:
Philip Hyunsu Cho
2020-07-25 13:46:19 -07:00
committed by GitHub
parent a4de2f68e4
commit 487ab0ce73
5 changed files with 79 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/data/data.cc
View File

@@ -833,9 +833,9 @@ uint64_t SparsePage::Push(const AdapterBatchT& batch, float missing, int nthread
uint64_t max_columns = 0;
// First-pass over the batch counting valid elements
size_t num_lines = batch.Size();
size_t batch_size = batch.Size();
#pragma omp parallel for schedule(static)
for (omp_ulong i = 0; i < static_cast<omp_ulong>(num_lines);
for (omp_ulong i = 0; i < static_cast<omp_ulong>(batch_size);
++i) { // NOLINT(*)
int tid = omp_get_thread_num();
auto line = batch.GetLine(i);
@@ -847,7 +847,7 @@ uint64_t SparsePage::Push(const AdapterBatchT& batch, float missing, int nthread
size_t key = element.row_idx - base_rowid;
// Adapter row index is absolute, here we want it relative to
// current page
CHECK_GE(key, builder_base_row_offset);
CHECK_GE(key, builder_base_row_offset);
builder.AddBudget(key, tid);
}
}
@@ -856,7 +856,7 @@ uint64_t SparsePage::Push(const AdapterBatchT& batch, float missing, int nthread
// Second pass over batch, placing elements in correct position
#pragma omp parallel for schedule(static)
for (omp_ulong i = 0; i < static_cast<omp_ulong>(num_lines);
for (omp_ulong i = 0; i < static_cast<omp_ulong>(batch_size);
++i) { // NOLINT(*)
int tid = omp_get_thread_num();
auto line = batch.GetLine(i);