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

Write ELLPACK pages to disk (#4879)

Browse Source 
* add ellpack source
* add batch param
* extract function to parse cache info
* construct ellpack info separately
* push batch to ellpack page
* write ellpack page.
* make sparse page source reusable
This commit is contained in:
Rong Ou
2019-10-22 20:44:32 -07:00
committed by Jiaming Yuan
parent 310fe60b35
commit 5b1715d97c
25 changed files with 935 additions and 408 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
tests/cpp/data/test_ellpack_page.cu
View File

@@ -17,15 +17,13 @@ TEST(EllpackPage, EmptyDMatrix) {
constexpr int kNRows = 0, kNCols = 0, kMaxBin = 256, kGpuBatchNRows = 64;
constexpr float kSparsity = 0;
auto dmat = *CreateDMatrix(kNRows, kNCols, kSparsity);
auto& page = *dmat->GetBatches<EllpackPage>().begin();
auto& page = *dmat->GetBatches<EllpackPage>({0, kMaxBin, kGpuBatchNRows}).begin();
auto impl = page.Impl();
impl->Init(0, kMaxBin, kGpuBatchNRows);
ASSERT_EQ(impl->ellpack_matrix.feature_segments.size(), 1);
ASSERT_EQ(impl->ellpack_matrix.min_fvalue.size(), 0);
ASSERT_EQ(impl->ellpack_matrix.gidx_fvalue_map.size(), 0);
ASSERT_EQ(impl->ellpack_matrix.row_stride, 0);
ASSERT_EQ(impl->ellpack_matrix.null_gidx_value, 0);
ASSERT_EQ(impl->n_bins, 0);
ASSERT_EQ(impl->matrix.info.feature_segments.size(), 1);
ASSERT_EQ(impl->matrix.info.min_fvalue.size(), 0);
ASSERT_EQ(impl->matrix.info.gidx_fvalue_map.size(), 0);
ASSERT_EQ(impl->matrix.info.row_stride, 0);
ASSERT_EQ(impl->matrix.info.n_bins, 0);
ASSERT_EQ(impl->gidx_buffer.size(), 4);
}
@@ -37,7 +35,7 @@ TEST(EllpackPage, BuildGidxDense) {
dh::CopyDeviceSpanToVector(&h_gidx_buffer, page->gidx_buffer);
common::CompressedIterator<uint32_t> gidx(h_gidx_buffer.data(), 25);
ASSERT_EQ(page->ellpack_matrix.row_stride, kNCols);
ASSERT_EQ(page->matrix.info.row_stride, kNCols);
std::vector<uint32_t> solution = {
0, 3, 8, 9, 14, 17, 20, 21,
@@ -70,7 +68,7 @@ TEST(EllpackPage, BuildGidxSparse) {
dh::CopyDeviceSpanToVector(&h_gidx_buffer, page->gidx_buffer);
common::CompressedIterator<uint32_t> gidx(h_gidx_buffer.data(), 25);
ASSERT_LE(page->ellpack_matrix.row_stride, 3);
ASSERT_LE(page->matrix.info.row_stride, 3);
// row_stride = 3, 16 rows, 48 entries for ELLPack
std::vector<uint32_t> solution = {
@@ -78,7 +76,7 @@ TEST(EllpackPage, BuildGidxSparse) {
24, 24, 24, 24, 24, 5, 24, 24, 0, 16, 24, 15, 24, 24, 24, 24,
24, 7, 14, 16, 4, 24, 24, 24, 24, 24, 9, 24, 24, 1, 24, 24
};
for (size_t i = 0; i < kNRows * page->ellpack_matrix.row_stride; ++i) {
for (size_t i = 0; i < kNRows * page->matrix.info.row_stride; ++i) {
ASSERT_EQ(solution[i], gidx[i]);
}
}

26
tests/cpp/data/test_sparse_page_dmatrix.cu Normal file
View File

@@ -0,0 +1,26 @@
// Copyright by Contributors
#include <dmlc/filesystem.h>
#include "../helpers.h"
namespace xgboost {
TEST(GPUSparsePageDMatrix, EllpackPage) {
dmlc::TemporaryDirectory tempdir;
const std::string tmp_file = tempdir.path + "/simple.libsvm";
CreateSimpleTestData(tmp_file);
DMatrix* dmat = DMatrix::Load(tmp_file + "#" + tmp_file + ".cache", true, false);
// Loop over the batches and assert the data is as expected
for (const auto& batch : dmat->GetBatches<EllpackPage>({0, 256, 64})) {
EXPECT_EQ(batch.Size(), dmat->Info().num_row_);
}
EXPECT_TRUE(FileExists(tmp_file + ".cache"));
EXPECT_TRUE(FileExists(tmp_file + ".cache.row.page"));
EXPECT_TRUE(FileExists(tmp_file + ".cache.ellpack.page"));
delete dmat;
}
} // namespace xgboost

6
tests/cpp/helpers.cc
View File

@@ -192,14 +192,14 @@ std::unique_ptr<DMatrix> CreateSparsePageDMatrix(
return dmat;
}
std::unique_ptr<DMatrix> CreateSparsePageDMatrixWithRC(size_t n_rows, size_t n_cols,
size_t page_size, bool deterministic) {
std::unique_ptr<DMatrix> CreateSparsePageDMatrixWithRC(
size_t n_rows, size_t n_cols, size_t page_size, bool deterministic,
const dmlc::TemporaryDirectory& tempdir) {
if (!n_rows || !n_cols) {
return nullptr;
}
// Create the svm file in a temp dir
dmlc::TemporaryDirectory tempdir;
const std::string tmp_file = tempdir.path + "/big.libsvm";
std::ofstream fo(tmp_file.c_str());

13
tests/cpp/helpers.h
View File

@@ -14,6 +14,7 @@
#include <gtest/gtest.h>
#include <dmlc/filesystem.h>
#include <xgboost/base.h>
#include <xgboost/objective.h>
#include <xgboost/metric.h>
@@ -199,8 +200,9 @@ std::unique_ptr<DMatrix> CreateSparsePageDMatrix(
*
* \return The new dmatrix.
*/
std::unique_ptr<DMatrix> CreateSparsePageDMatrixWithRC(size_t n_rows, size_t n_cols,
size_t page_size, bool deterministic);
std::unique_ptr<DMatrix> CreateSparsePageDMatrixWithRC(
size_t n_rows, size_t n_cols, size_t page_size, bool deterministic,
const dmlc::TemporaryDirectory& tempdir = dmlc::TemporaryDirectory());
gbm::GBTreeModel CreateTestModel();
@@ -247,16 +249,15 @@ inline std::unique_ptr<EllpackPageImpl> BuildEllpackPage(
0.26f, 0.71f, 1.83f});
cmat.SetMins({0.1f, 0.2f, 0.3f, 0.1f, 0.2f, 0.3f, 0.2f, 0.2f});
auto is_dense = (*dmat)->Info().num_nonzero_ ==
(*dmat)->Info().num_row_ * (*dmat)->Info().num_col_;
size_t row_stride = 0;
const auto &offset_vec = batch.offset.ConstHostVector();
for (size_t i = 1; i < offset_vec.size(); ++i) {
row_stride = std::max(row_stride, offset_vec[i] - offset_vec[i-1]);
}
auto page = std::unique_ptr<EllpackPageImpl>(new EllpackPageImpl(dmat->get()));
page->InitCompressedData(0, cmat, row_stride, is_dense);
auto page = std::unique_ptr<EllpackPageImpl>(new EllpackPageImpl(dmat->get(), {0, 256, 0}));
page->InitInfo(0, (*dmat)->IsDense(), row_stride, cmat);
page->InitCompressedData(0, n_rows);
page->CreateHistIndices(0, batch, RowStateOnDevice(batch.Size(), batch.Size()));
delete dmat;

24
tests/cpp/tree/test_gpu_hist.cu
View File

@@ -2,6 +2,7 @@
* Copyright 2017-2019 XGBoost contributors
*/
#include <thrust/device_vector.h>
#include <dmlc/filesystem.h>
#include <xgboost/base.h>
#include <random>
#include <string>
@@ -207,14 +208,14 @@ TEST(GpuHist, EvaluateSplits) {
// Copy cut matrix to device.
maker.ba.Allocate(0,
&(page->ellpack_matrix.feature_segments), cmat.Ptrs().size(),
&(page->ellpack_matrix.min_fvalue), cmat.MinValues().size(),
&(page->ellpack_matrix.gidx_fvalue_map), 24,
&(page->matrix.info.feature_segments), cmat.Ptrs().size(),
&(page->matrix.info.min_fvalue), cmat.MinValues().size(),
&(page->matrix.info.gidx_fvalue_map), 24,
&(maker.monotone_constraints), kNCols);
dh::CopyVectorToDeviceSpan(page->ellpack_matrix.feature_segments, cmat.Ptrs());
dh::CopyVectorToDeviceSpan(page->ellpack_matrix.gidx_fvalue_map, cmat.Values());
dh::CopyVectorToDeviceSpan(page->matrix.info.feature_segments, cmat.Ptrs());
dh::CopyVectorToDeviceSpan(page->matrix.info.gidx_fvalue_map, cmat.Values());
dh::CopyVectorToDeviceSpan(maker.monotone_constraints, param.monotone_constraints);
dh::CopyVectorToDeviceSpan(page->ellpack_matrix.min_fvalue, cmat.MinValues());
dh::CopyVectorToDeviceSpan(page->matrix.info.min_fvalue, cmat.MinValues());
// Initialize GPUHistMakerDevice::hist
maker.hist.Init(0, (max_bins - 1) * kNCols);
@@ -265,8 +266,10 @@ void TestHistogramIndexImpl() {
tree::GPUHistMakerSpecialised<GradientPairPrecise> hist_maker, hist_maker_ext;
std::unique_ptr<DMatrix> hist_maker_dmat(
CreateSparsePageDMatrixWithRC(kNRows, kNCols, 0, true));
dmlc::TemporaryDirectory tempdir;
std::unique_ptr<DMatrix> hist_maker_ext_dmat(
CreateSparsePageDMatrixWithRC(kNRows, kNCols, 128UL, true));
CreateSparsePageDMatrixWithRC(kNRows, kNCols, 128UL, true, tempdir));
std::vector<std::pair<std::string, std::string>> training_params = {
{"max_depth", "10"},
@@ -275,22 +278,21 @@ void TestHistogramIndexImpl() {
GenericParameter generic_param(CreateEmptyGenericParam(0));
hist_maker.Configure(training_params, &generic_param);
hist_maker.InitDataOnce(hist_maker_dmat.get());
hist_maker_ext.Configure(training_params, &generic_param);
hist_maker_ext.InitDataOnce(hist_maker_ext_dmat.get());
// Extract the device maker from the histogram makers and from that its compressed
// histogram index
const auto &maker = hist_maker.maker_;
const auto &maker = hist_maker.maker;
std::vector<common::CompressedByteT> h_gidx_buffer(maker->page->gidx_buffer.size());
dh::CopyDeviceSpanToVector(&h_gidx_buffer, maker->page->gidx_buffer);
const auto &maker_ext = hist_maker_ext.maker_;
const auto &maker_ext = hist_maker_ext.maker;
std::vector<common::CompressedByteT> h_gidx_buffer_ext(maker_ext->page->gidx_buffer.size());
dh::CopyDeviceSpanToVector(&h_gidx_buffer_ext, maker_ext->page->gidx_buffer);
ASSERT_EQ(maker->page->n_bins, maker_ext->page->n_bins);
ASSERT_EQ(maker->page->matrix.info.n_bins, maker_ext->page->matrix.info.n_bins);
ASSERT_EQ(maker->page->gidx_buffer.size(), maker_ext->page->gidx_buffer.size());
ASSERT_EQ(h_gidx_buffer, h_gidx_buffer_ext);