pitrou commented on code in PR #46926:
URL: https://github.com/apache/arrow/pull/46926#discussion_r3450989626
##########
cpp/src/arrow/compute/kernels/vector_select_k.cc:
##########
@@ -205,40 +317,94 @@ class ChunkedArraySelector : public TypeVisitor {
if (k_ > chunked_array_.length()) {
k_ = chunked_array_.length();
}
+
+ ARROW_ASSIGN_OR_RAISE(auto take_indices,
+ MakeMutableUInt64Array(k_, ctx_->memory_pool()));
+ auto* output_begin = take_indices->GetMutableValues<uint64_t>(1);
+
+ std::vector<std::shared_ptr<ArrayType>> chunks_holder;
+ chunks_holder.reserve(num_chunks);
+ std::vector<PartitionResultByNullLikeness> partitions_by_chunk;
+ partitions_by_chunk.reserve(num_chunks);
+ std::vector<std::vector<uint64_t>> indices_by_chunk;
+ indices_by_chunk.reserve(num_chunks);
+
+ int64_t null_count = 0;
+ int64_t nan_count = 0;
+ int64_t non_null_like_count = 0;
+
+ for (const auto& chunk : physical_chunks_) {
+ if (chunk->length() == 0) continue;
+ chunks_holder.emplace_back(std::make_shared<ArrayType>(chunk->data()));
+ ArrayType& arr = *chunks_holder.back();
+
+ auto& indices = indices_by_chunk.emplace_back();
+ indices.resize(arr.length());
+ uint64_t* indices_begin = indices.data();
+ uint64_t* indices_end = indices_begin + indices.size();
+ std::iota(indices_begin, indices_end, 0);
+
+ partitions_by_chunk.emplace_back(
+ PartitionNullsAndNans<ArrayType, NonStablePartitioner>(
+ indices_begin, indices_end, arr, 0, null_placement_));
+
+ null_count += partitions_by_chunk.back().null_range.size();
+ nan_count += partitions_by_chunk.back().nan_range.size();
+ non_null_like_count +=
partitions_by_chunk.back().non_null_like_range.size();
+ }
+ DCHECK_EQ(non_null_like_count + null_count + nan_count,
chunked_array_.length());
+
+ auto output = CalculateOutputRangesByNullLikeness(non_null_like_count,
nan_count,
+ null_count,
null_placement_,
+ {output_begin,
output_begin + k_});
+
+ // Now we can independently fill the output with non_null, nan and null
items.
+ // For non_null, we do a heap_sort, the others can just be copied until
+ // nan_taken == output.nan_range.size() and
+ // null_taken == output.null_range.size() respectively
+ size_t nan_taken = 0;
+ size_t null_taken = 0;
+
std::function<bool(const HeapItem&, const HeapItem&)> cmp;
SelectKComparator<sort_order> comparator;
-
cmp = [&comparator](const HeapItem& left, const HeapItem& right) -> bool {
const auto lval = GetView::LogicalValue(left.array->GetView(left.index));
const auto rval =
GetView::LogicalValue(right.array->GetView(right.index));
return comparator(lval, rval);
};
using HeapContainer =
std::priority_queue<HeapItem, std::vector<HeapItem>, decltype(cmp)>;
-
HeapContainer heap(cmp);
- std::vector<std::shared_ptr<ArrayType>> chunks_holder;
- uint64_t offset = 0;
- for (const auto& chunk : physical_chunks_) {
- if (chunk->length() == 0) continue;
- chunks_holder.emplace_back(std::make_shared<ArrayType>(chunk->data()));
- ArrayType& arr = *chunks_holder[chunks_holder.size() - 1];
- std::vector<uint64_t> indices(arr.length());
- uint64_t* indices_begin = indices.data();
- uint64_t* indices_end = indices_begin + indices.size();
- std::iota(indices_begin, indices_end, 0);
+ DCHECK_EQ(chunks_holder.size(), partitions_by_chunk.size());
+ uint64_t offset = 0;
+ for (size_t chunk_id = 0; chunk_id < chunks_holder.size(); ++chunk_id) {
+ ArrayType& arr = *chunks_holder[chunk_id];
Review Comment:
Make this `const` too.
##########
cpp/src/arrow/compute/kernels/vector_select_k.cc:
##########
@@ -205,40 +317,94 @@ class ChunkedArraySelector : public TypeVisitor {
if (k_ > chunked_array_.length()) {
k_ = chunked_array_.length();
}
+
+ ARROW_ASSIGN_OR_RAISE(auto take_indices,
+ MakeMutableUInt64Array(k_, ctx_->memory_pool()));
+ auto* output_begin = take_indices->GetMutableValues<uint64_t>(1);
+
+ std::vector<std::shared_ptr<ArrayType>> chunks_holder;
+ chunks_holder.reserve(num_chunks);
+ std::vector<PartitionResultByNullLikeness> partitions_by_chunk;
+ partitions_by_chunk.reserve(num_chunks);
+ std::vector<std::vector<uint64_t>> indices_by_chunk;
+ indices_by_chunk.reserve(num_chunks);
+
+ int64_t null_count = 0;
+ int64_t nan_count = 0;
+ int64_t non_null_like_count = 0;
+
+ for (const auto& chunk : physical_chunks_) {
+ if (chunk->length() == 0) continue;
+ chunks_holder.emplace_back(std::make_shared<ArrayType>(chunk->data()));
+ ArrayType& arr = *chunks_holder.back();
Review Comment:
Let's please make this `const`?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
