Add number of columns to native data iterator. (#5202)

* Change native data iter into an adapter.
This commit is contained in:
Jiaming Yuan
2020-02-25 23:42:01 +08:00
committed by GitHub
parent e0509b3307
commit f2b8cd2922
11 changed files with 244 additions and 156 deletions

View File

@@ -1,18 +1,24 @@
// Copyright (c) 2019 by Contributors
#include <gtest/gtest.h>
#include <type_traits>
#include <utility>
#include <xgboost/data.h>
#include "../../../src/data/adapter.h"
#include "../../../src/data/simple_dmatrix.h"
#include "../../../src/common/timer.h"
#include "../helpers.h"
using namespace xgboost; // NOLINT
TEST(adapter, CSRAdapter) {
#include "xgboost/base.h"
#include "xgboost/c_api.h"
namespace xgboost {
TEST(Adapter, CSRAdapter) {
int n = 2;
std::vector<float> data = {1, 2, 3, 4, 5};
std::vector<unsigned> feature_idx = {0, 1, 0, 1, 1};
std::vector<size_t> row_ptr = {0, 2, 4, 5};
data::CSRAdapter adapter(row_ptr.data(), feature_idx.data(), data.data(),
row_ptr.size() - 1, data.size(), n);
row_ptr.size() - 1, data.size(), n);
adapter.Next();
auto & batch = adapter.Value();
auto line0 = batch.GetLine(0);
@@ -28,7 +34,7 @@ TEST(adapter, CSRAdapter) {
EXPECT_EQ(line2 .GetElement(0).column_idx, 1);
}
TEST(adapter, CSCAdapterColsMoreThanRows) {
TEST(Adapter, CSCAdapterColsMoreThanRows) {
std::vector<float> data = {1, 2, 3, 4, 5, 6, 7, 8};
std::vector<unsigned> row_idx = {0, 1, 0, 1, 0, 1, 0, 1};
std::vector<size_t> col_ptr = {0, 2, 4, 6, 8};
@@ -88,3 +94,67 @@ TEST(c_api, DMatrixSliceAdapterFromSimpleDMatrix) {
delete pp_dmat;
}
// A mock for JVM data iterator.
class DataIterForTest {
std::vector<float> data_ {1, 2, 3, 4, 5};
std::vector<std::remove_pointer<decltype(std::declval<XGBoostBatchCSR>().index)>::type>
feature_idx_ {0, 1, 0, 1, 1};
std::vector<std::remove_pointer<decltype(std::declval<XGBoostBatchCSR>().offset)>::type>
row_ptr_ {0, 2, 4, 5};
size_t iter_ {0};
public:
size_t static constexpr kCols { 13 }; // Test for having some missing columns
XGBoostBatchCSR Next() {
for (auto& v : data_) {
v += iter_;
}
XGBoostBatchCSR batch;
batch.columns = 2;
batch.offset = dmlc::BeginPtr(row_ptr_);
batch.index = dmlc::BeginPtr(feature_idx_);
batch.value = dmlc::BeginPtr(data_);
batch.size = 3;
batch.label = nullptr;
batch.weight = nullptr;
iter_++;
return batch;
}
size_t Iter() const { return iter_; }
};
size_t constexpr DataIterForTest::kCols;
int SetDataNextForTest(DataIterHandle data_handle,
XGBCallbackSetData *set_function,
DataHolderHandle set_function_handle) {
size_t constexpr kIters { 2 };
auto iter = static_cast<DataIterForTest *>(data_handle);
if (iter->Iter() < kIters) {
auto batch = iter->Next();
batch.columns = DataIterForTest::kCols;
set_function(set_function_handle, batch);
return 1;
} else {
return 0; // stoping condition
}
}
TEST(Adapter, IteratorAdaper) {
DataIterForTest iter;
data::IteratorAdapter adapter{&iter, SetDataNextForTest};
constexpr size_t kRows { 6 };
std::unique_ptr<DMatrix> data {
DMatrix::Create(&adapter, std::numeric_limits<float>::quiet_NaN(), 1)
};
ASSERT_EQ(data->Info().num_col_, DataIterForTest::kCols);
ASSERT_EQ(data->Info().num_row_, kRows);
}
} // namespace xgboost

View File

@@ -5,6 +5,7 @@
#include "../../../src/data/adapter.h"
#include "../helpers.h"
#include "xgboost/base.h"
using namespace xgboost; // NOLINT
@@ -185,10 +186,8 @@ TEST(SimpleDMatrix, FromFile) {
CreateBigTestData(filename, 3 * 5);
std::unique_ptr<dmlc::Parser<uint32_t>> parser(
dmlc::Parser<uint32_t>::Create(filename.c_str(), 0, 1, "auto"));
data::FileAdapter adapter(parser.get());
data::SimpleDMatrix dmat(&adapter, std::numeric_limits<float>::quiet_NaN(),
1);
for (auto &batch : dmat.GetBatches<SparsePage>()) {
auto verify_batch = [](SparsePage const &batch) {
EXPECT_EQ(batch.Size(), 5);
EXPECT_EQ(batch.offset.HostVector(),
std::vector<bst_row_t>({0, 3, 6, 9, 12, 15}));
@@ -205,6 +204,16 @@ TEST(SimpleDMatrix, FromFile) {
EXPECT_EQ(batch[i][2].index, 4);
}
}
};
constexpr bst_feature_t kCols = 5;
data::FileAdapter adapter(parser.get());
data::SimpleDMatrix dmat(&adapter, std::numeric_limits<float>::quiet_NaN(),
1);
ASSERT_EQ(dmat.Info().num_col_, kCols);
for (auto &batch : dmat.GetBatches<SparsePage>()) {
verify_batch(batch);
}
}

View File

@@ -263,8 +263,10 @@ TEST(SparsePageDMatrix, FromFile) {
data::FileAdapter adapter(parser.get());
dmlc::TemporaryDirectory tempdir;
const std::string tmp_file = tempdir.path + "/simple.libsvm";
data::SparsePageDMatrix dmat(
&adapter, std::numeric_limits<float>::quiet_NaN(), -1, tmp_file, 1);
ASSERT_EQ(dmat.Info().num_col_, 5);
for (auto &batch : dmat.GetBatches<SparsePage>()) {
std::vector<bst_row_t> expected_offset(batch.Size() + 1);