Cleanup.

trivialfis · Jun 13, 2023 · c195db1 · c195db1
1 parent bcf4cdb
commit c195db1
Show file tree

Hide file tree

Showing 5 changed files with 47 additions and 7 deletions.
diff --git a/src/common/io.cc b/src/common/io.cc
@@ -212,8 +212,6 @@ char* PrivateMmapStream::Open(std::string path, bool read_only, std::size_t offs
   char* ptr{nullptr};
   auto view_start = offset / GetPageSize() * GetPageSize();
   auto view_size = length + (offset - view_start);
-  std::cout << "offset:" << offset << ", length:" << length << ", start:" << view_start << ", size:" << view_size << std::endl;
-  // std::cout << view_start << " size: " << view_size << std::endl;
 #if defined(__linux__) || defined(__GLIBC__)
   int prot{PROT_READ};
   if (!read_only) {

diff --git a/src/data/sparse_page_source.h b/src/data/sparse_page_source.h
@@ -115,7 +115,7 @@ class SparsePageSourceImpl : public BatchIteratorImpl<S> {
     }
     // An heuristic for number of pre-fetched batches.  We can make it part of BatchParam
     // to let user adjust number of pre-fetched batches when needed.
-    uint32_t constexpr kPreFetch = 1;
+    uint32_t constexpr kPreFetch = 4;
 
     size_t n_prefetch_batches = std::min(kPreFetch, n_batches_);
     CHECK_GT(n_prefetch_batches, 0) << "total batches:" << n_batches_;
@@ -132,7 +132,7 @@ class SparsePageSourceImpl : public BatchIteratorImpl<S> {
       CHECK_LT(fetch_it, cache_info_->offset.size());
       ring_->at(fetch_it) = std::async(std::launch::async, [fetch_it, self, this]() {
         auto page = std::make_shared<S>();
-//        this->exec_.Run([&] {
+        this->exec_.Run([&] {
           common::Timer timer;
           timer.Start();
           std::unique_ptr<SparsePageFormat<S>> fmt{CreatePageFormat<S>("raw")};
@@ -144,7 +144,7 @@ class SparsePageSourceImpl : public BatchIteratorImpl<S> {
           auto fi = std::make_unique<common::PrivateMmapStream>(n, true, offset, length);
           CHECK(fmt->Read(page.get(), fi.get()));
           LOG(INFO) << "Read a page in " << timer.ElapsedSeconds() << " seconds.";
- //       });
+        });
         return page;
       });
     }
@@ -174,7 +174,6 @@ class SparsePageSourceImpl : public BatchIteratorImpl<S> {
     }
 
     auto bytes = fmt->Write(*page_, fo.get());
-    std::cout << "wrote: " << bytes << std::endl;
 
     timer.Stop();
     LOG(INFO) << static_cast<double>(bytes) / 1024.0 / 1024.0 << " MB written in "

diff --git a/tests/cpp/data/test_sparse_page_raw_format.cc b/tests/cpp/data/test_sparse_page_raw_format.cc
@@ -59,5 +59,40 @@ TEST(SparsePageRawFormat, CSCPage) {
 TEST(SparsePageRawFormat, SortedCSCPage) {
   TestSparsePageRawFormat<SortedCSCPage>();
 }
+
+TEST(Debug, WritePage) {
+  std::string path {"testfile"};
+  std::unique_ptr<SparsePageFormat<SparsePage>> fmt{CreatePageFormat<SparsePage>("raw")};
+  auto Xy = RandomDataGenerator{ 8192 * 8, 12, 0.0 }.GenerateDMatrix();
+  {
+    std::unique_ptr<dmlc::Stream> fo{dmlc::Stream::Create(path.c_str(), "w")};
+    for (auto const& page : Xy->GetBatches<SparsePage>()) {
+      std::cout << "back:" << page.offset.HostVector().back() << std::endl;
+      fmt->Write(page, fo.get());
+    }
+  }
+  {
+    std::unique_ptr<dmlc::Stream> fo{dmlc::Stream::Create(path.c_str(), "ab")};
+    for (auto const& page : Xy->GetBatches<SparsePage>()) {
+      std::cout << "back:" << page.offset.HostVector().back() << std::endl;
+      fmt->Write(page, fo.get());
+    }
+  }
+
+  {
+    std::unique_ptr<SparsePageFormat<SparsePage>> fmt{CreatePageFormat<SparsePage>("raw")};
+    std::unique_ptr<dmlc::SeekStream> fi{dmlc::SeekStream::CreateForRead(path.c_str())};
+    {
+      SparsePage page;
+      fmt->Read(&page, fi.get());
+      std::cout << "back:" << page.offset.HostVector().back() << std::endl;
+    }
+    {
+      SparsePage page;
+      fmt->Read(&page, fi.get());
+      std::cout << "back:" << page.offset.HostVector().back() << std::endl;
+    }
+  }
+}
 }  // namespace data
 }  // namespace xgboost
diff --git a/tests/cpp/test_main.cc b/tests/cpp/test_main.cc
@@ -11,6 +11,7 @@
 int main(int argc, char ** argv) {
   xgboost::Args args {{"verbosity", "2"}};
   xgboost::ConsoleLogger::Configure(args);
+
   testing::InitGoogleTest(&argc, argv);
   testing::FLAGS_gtest_death_test_style = "threadsafe";
   auto rmm_alloc = xgboost::SetUpRMMResourceForCppTests(argc, argv);

diff --git a/tests/cpp/tree/hist/test_histogram.cc b/tests/cpp/tree/hist/test_histogram.cc
@@ -477,8 +477,15 @@ TEST(CPUHistogram, ExternalMemory) {
   int32_t constexpr kBins = 256;
   Context ctx;
 
-  std::cout << "l:" << __LINE__ << std::endl;
   TestHistogramExternalMemory(&ctx, BatchParam{kBins, common::Span<float>{}, false}, true, false);
+  TestHistogramExternalMemory(&ctx, BatchParam{kBins, common::Span<float>{}, false}, true, true);
+
+  float sparse_thresh{0.5};
+  TestHistogramExternalMemory(&ctx, {kBins, sparse_thresh}, false, false);
+  TestHistogramExternalMemory(&ctx, {kBins, sparse_thresh}, false, true);
+  sparse_thresh = std::numeric_limits<float>::quiet_NaN();
+  TestHistogramExternalMemory(&ctx, {kBins, sparse_thresh}, false, false);
+  TestHistogramExternalMemory(&ctx, {kBins, sparse_thresh}, false, true);
 }
 }  // namespace tree
 }  // namespace xgboost