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

Rewrite sparse dmatrix using callbacks. (#7092)

Browse Source 
- Reduce dependency on dmlc parsers and provide an interface for users to load data by themselves.
- Remove use of threaded iterator and IO queue.
- Remove `page_size`.
- Make sure the number of pages in memory is bounded.
- Make sure the cache can not be violated.
- Provide an interface for internal algorithms to process data asynchronously.
This commit is contained in:
Jiaming Yuan
2021-07-16 12:33:31 +08:00
committed by GitHub
parent 2f524e9f41
commit bd1f3a38f0
51 changed files with 1445 additions and 1391 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/cpp/data/test_ellpack_page.cu
View File

@@ -142,7 +142,7 @@ TEST(EllpackPage, Copy) {
dmlc::TemporaryDirectory tmpdir;
std::unique_ptr<DMatrix>
dmat(CreateSparsePageDMatrixWithRC(kRows, kCols, kPageSize, true, tmpdir));
BatchParam param{0, 256, kPageSize};
BatchParam param{0, 256};
auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
// Create an empty result page.
@@ -188,7 +188,7 @@ TEST(EllpackPage, Compact) {
dmlc::TemporaryDirectory tmpdir;
std::unique_ptr<DMatrix>
dmat(CreateSparsePageDMatrixWithRC(kRows, kCols, kPageSize, true, tmpdir));
BatchParam param{0, 256, kPageSize};
BatchParam param{0, 256};
auto page = (*dmat->GetBatches<EllpackPage>(param).begin()).Impl();
// Create an empty result page.
@@ -212,7 +212,7 @@ TEST(EllpackPage, Compact) {
std::vector<bst_float> row_result(kCols);
for (auto& page : dmat->GetBatches<EllpackPage>(param)) {
auto impl = page.Impl();
EXPECT_EQ(impl->base_rowid, current_row);
ASSERT_EQ(impl->base_rowid, current_row);
for (size_t i = 0; i < impl->Size(); i++) {
size_t compacted_row = row_indexes_h[current_row];