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

Define bin type. (#7850)

Browse Source
This commit is contained in:
Jiaming Yuan
2022-04-29 19:41:39 +08:00
committed by GitHub
parent f7db16add1
commit 288c52596c
4 changed files with 25 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
src/common/column_matrix.h
View File

@@ -34,8 +34,8 @@ class Column {
public:
static constexpr int32_t kMissingId = -1;
Column(ColumnType type, common::Span<const BinIdxType> index, const uint32_t index_base)
: type_(type), index_(index), index_base_(index_base) {}
Column(ColumnType type, common::Span<const BinIdxType> index, const bst_bin_t index_base)
: type_(type), index_(index), index_base_{index_base} {}
virtual ~Column() = default;
@@ -60,19 +60,19 @@ class Column {
/* bin indexes in range [0, max_bins - 1] */
common::Span<const BinIdxType> index_;
/* bin index offset for specific feature */
const uint32_t index_base_;
bst_bin_t const index_base_;
};
template <typename BinIdxType>
class SparseColumn : public Column<BinIdxType> {
public:
SparseColumn(ColumnType type, common::Span<const BinIdxType> index, uint32_t index_base,
SparseColumn(ColumnType type, common::Span<const BinIdxType> index, bst_bin_t index_base,
common::Span<const size_t> row_ind)
: Column<BinIdxType>(type, index, index_base), row_ind_(row_ind) {}
const size_t* GetRowData() const { return row_ind_.data(); }
int32_t GetBinIdx(size_t rid, size_t* state) const {
bst_bin_t GetBinIdx(size_t rid, size_t* state) const {
const size_t column_size = this->Size();
if (!((*state) < column_size)) {
return this->kMissingId;

24
src/common/hist_util.h
View File

@@ -40,8 +40,6 @@ class HistogramCuts {
float max_cat_{-1.0f};
protected:
using BinIdx = uint32_t;
void Swap(HistogramCuts&& that) noexcept(true) {
std::swap(cut_values_, that.cut_values_);
std::swap(cut_ptrs_, that.cut_ptrs_);
@@ -110,31 +108,31 @@ class HistogramCuts {
// Return the index of a cut point that is strictly greater than the input
// value, or the last available index if none exists
BinIdx SearchBin(float value, bst_feature_t column_id, std::vector<uint32_t> const& ptrs,
std::vector<float> const& values) const {
bst_bin_t SearchBin(float value, bst_feature_t column_id, std::vector<uint32_t> const& ptrs,
std::vector<float> const& values) const {
auto end = ptrs[column_id + 1];
auto beg = ptrs[column_id];
auto it = std::upper_bound(values.cbegin() + beg, values.cbegin() + end, value);
BinIdx idx = it - values.cbegin();
bst_bin_t idx = it - values.cbegin();
idx -= !!(idx == end);
return idx;
}
BinIdx SearchBin(float value, bst_feature_t column_id) const {
bst_bin_t SearchBin(float value, bst_feature_t column_id) const {
return this->SearchBin(value, column_id, Ptrs(), Values());
}
/**
* \brief Search the bin index for numerical feature.
*/
BinIdx SearchBin(Entry const& e) const {
bst_bin_t SearchBin(Entry const& e) const {
return SearchBin(e.fvalue, e.index);
}
/**
* \brief Search the bin index for categorical feature.
*/
BinIdx SearchCatBin(Entry const &e) const {
bst_bin_t SearchCatBin(Entry const &e) const {
auto const &ptrs = this->Ptrs();
auto const &vals = this->Values();
auto end = ptrs.at(e.index + 1) + vals.cbegin();
@@ -296,10 +294,10 @@ struct Index {
};
template <typename GradientIndex>
int32_t XGBOOST_HOST_DEV_INLINE BinarySearchBin(size_t begin, size_t end,
GradientIndex const &data,
uint32_t const fidx_begin,
uint32_t const fidx_end) {
bst_bin_t XGBOOST_HOST_DEV_INLINE BinarySearchBin(size_t begin, size_t end,
GradientIndex const& data,
uint32_t const fidx_begin,
uint32_t const fidx_end) {
size_t previous_middle = std::numeric_limits<size_t>::max();
while (end != begin) {
size_t middle = begin + (end - begin) / 2;
@@ -324,8 +322,6 @@ int32_t XGBOOST_HOST_DEV_INLINE BinarySearchBin(size_t begin, size_t end,
return -1;
}
class ColumnMatrix;
template<typename GradientSumT>
using GHistRow = Span<xgboost::detail::GradientPairInternal<GradientSumT> >;