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

skip missing lookup if nothing is missing in CPU hist partition kernel. (#5644)

Browse Source 
* [xgboost] skip missing lookup if nothing is missing
This commit is contained in:
Oleksandr Kuvshynov
2020-05-11 19:50:08 -07:00
committed by GitHub
parent 9ad40901a8
commit 4e64e2ef8e
3 changed files with 123 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
src/common/column_matrix.h
View File

@@ -154,6 +154,7 @@ class ColumnMatrix {
index_base_ = const_cast<uint32_t*>(gmat.cut.Ptrs().data());
const bool noMissingValues = NoMissingValues(gmat.row_ptr[nrow], nrow, nfeature);
any_missing_ = !noMissingValues;
if (noMissingValues) {
missing_flags_.resize(feature_offsets_[nfeature], false);
@@ -311,11 +312,18 @@ class ColumnMatrix {
const BinTypeSize GetTypeSize() const {
return bins_type_size_;
}
// This is just an utility function
const bool NoMissingValues(const size_t n_elements,
const size_t n_row, const size_t n_features) {
return n_elements == n_features * n_row;
}
// And this returns part of state
const bool AnyMissing() const {
return any_missing_;
}
private:
std::vector<uint8_t> index_;
@@ -329,6 +337,7 @@ class ColumnMatrix {
uint32_t* index_base_;
std::vector<bool> missing_flags_;
BinTypeSize bins_type_size_;
bool any_missing_;
};
} // namespace common

41
src/tree/updater_quantile_hist.cc
View File

@@ -826,7 +826,7 @@ void QuantileHistMaker::Builder::EvaluateSplits(const std::vector<ExpandEntry>&
// on comparison of indexes values (idx_span) and split point (split_cond)
// Handle dense columns
// Analog of std::stable_partition, but in no-inplace manner
template <bool default_left, typename BinIdxType>
template <bool default_left, bool any_missing, typename BinIdxType>
inline std::pair<size_t, size_t> PartitionDenseKernel(const common::DenseColumn<BinIdxType>& column,
common::Span<const size_t> rid_span, const int32_t split_cond,
common::Span<size_t> left_part, common::Span<size_t> right_part) {
@@ -837,14 +837,24 @@ inline std::pair<size_t, size_t> PartitionDenseKernel(const common::DenseColumn<
size_t nleft_elems = 0;
size_t nright_elems = 0;
for (auto rid : rid_span) {
if (column.IsMissing(rid)) {
if (default_left) {
p_left_part[nleft_elems++] = rid;
if (any_missing) {
for (auto rid : rid_span) {
if (column.IsMissing(rid)) {
if (default_left) {
p_left_part[nleft_elems++] = rid;
} else {
p_right_part[nright_elems++] = rid;
}
} else {
p_right_part[nright_elems++] = rid;
if ((static_cast<int32_t>(idx[rid]) + offset) <= split_cond) {
p_left_part[nleft_elems++] = rid;
} else {
p_right_part[nright_elems++] = rid;
}
}
} else {
}
} else {
for (auto rid : rid_span) {
if ((static_cast<int32_t>(idx[rid]) + offset) <= split_cond) {
p_left_part[nleft_elems++] = rid;
} else {
@@ -919,6 +929,7 @@ void QuantileHistMaker::Builder::PartitionKernel(
const size_t node_in_set, const size_t nid, common::Range1d range,
const int32_t split_cond, const ColumnMatrix& column_matrix, const RegTree& tree) {
const size_t* rid = row_set_collection_[nid].begin;
common::Span<const size_t> rid_span(rid + range.begin(), rid + range.end());
common::Span<size_t> left = partition_builder_.GetLeftBuffer(node_in_set,
range.begin(), range.end());
@@ -934,9 +945,21 @@ void QuantileHistMaker::Builder::PartitionKernel(
const common::DenseColumn<BinIdxType>& column =
static_cast<const common::DenseColumn<BinIdxType>& >(*(column_ptr.get()));
if (default_left) {
child_nodes_sizes = PartitionDenseKernel<true>(column, rid_span, split_cond, left, right);
if (column_matrix.AnyMissing()) {
child_nodes_sizes = PartitionDenseKernel<true, true>(column, rid_span, split_cond,
left, right);
} else {
child_nodes_sizes = PartitionDenseKernel<true, false>(column, rid_span, split_cond,
left, right);
}
} else {
child_nodes_sizes = PartitionDenseKernel<false>(column, rid_span, split_cond, left, right);
if (column_matrix.AnyMissing()) {
child_nodes_sizes = PartitionDenseKernel<false, true>(column, rid_span, split_cond,
left, right);
} else {
child_nodes_sizes = PartitionDenseKernel<false, false>(column, rid_span, split_cond,
left, right);
}
}
} else {
const common::SparseColumn<BinIdxType>& column