rtpsw commented on code in PR #34311:
URL: https://github.com/apache/arrow/pull/34311#discussion_r1125647820
##########
cpp/src/arrow/compute/exec/aggregate_node.cc:
##########
@@ -326,10 +437,22 @@ class ScalarAggregateNode : public ExecNode, public
TracedNode {
}
private:
- Status Finish() {
- auto scope = TraceFinish();
+ Status ResetAggregates() {
+ auto exec_ctx = plan()->query_context()->exec_context();
+ for (size_t i = 0; i < kernels_.size(); ++i) {
+ const std::vector<TypeHolder>& in_types = in_typesets_[i];
+ states_[i].resize(plan()->query_context()->max_concurrency());
+ KernelContext kernel_ctx{exec_ctx};
+ RETURN_NOT_OK(Kernel::InitAll(
+ &kernel_ctx, KernelInitArgs{kernels_[i], in_types,
aggs_[i].options.get()},
+ &states_[i]));
+ }
+ return Status::OK();
+ }
+
+ Status OutputResult(bool is_last = false) {
ExecBatch batch{{}, 1};
- batch.values.resize(kernels_.size());
+ batch.values.resize(kernels_.size() + segment_field_ids_.size());
Review Comment:
This code is in `ScalarAggregateNode`, which does not have keys but may have
segment-keys. The output batch for this class has two sets of columns, one set
of kernel/aggregate columns and a second set of segment-key columns. The
allocation here is for both sets of columns.
--
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]
