Skip to content

Lazily initialize streaming groupby aggregation kinds #23850

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
rapids-bot merged 1 commit into from
Aug 28, 2026
Merged

Lazily initialize streaming groupby aggregation kinds #23850

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion cpp/src/groupby/streaming_groupby/aggregate.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,7 +60,7 @@ void streaming_groupby::impl::do_aggregate(table_view const& data, cuda::stream_
cuda::counting_iterator<int64_t>(0),
static_cast<int64_t>(batch_size) * num_agg_cols,
detail::hash::compute_single_pass_aggs_dense_output_fn{
result.target_indices.begin(), _d_agg_kinds.data(), *d_values, *_d_agg_results});
result.target_indices.begin(), _d_agg_kinds->data(), *d_values, *_d_agg_results});
}

} // namespace cudf::groupby
2 changes: 1 addition & 1 deletion cpp/src/groupby/streaming_groupby/common.cuh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -296,7 +296,7 @@ struct streaming_groupby::impl {
*/
std::unique_ptr<mutable_table_device_view, void (*)(mutable_table_device_view*)> _d_agg_results;
std::vector<size_type> _value_col_indices;
rmm::device_uvector<aggregation::Kind> _d_agg_kinds;
std::unique_ptr<rmm::device_uvector<aggregation::Kind>> _d_agg_kinds;

std::unique_ptr<streaming_set_t> _key_set;

Expand Down
4 changes: 2 additions & 2 deletions cpp/src/groupby/streaming_groupby/impl.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,7 +89,6 @@ streaming_groupby::impl::impl(host_span<size_type const> key_indices,
: _max_distinct_keys{max_distinct_keys},
_null_handling{null_handling},
_mr{std::move(mr)},
_d_agg_kinds{0, cuda::stream_ref{cudaStreamLegacy}, cudf::get_current_device_resource_ref()},
_d_agg_results{nullptr, +[](mutable_table_device_view*) {}}
{
CUDF_EXPECTS(max_distinct_keys > 0, "max_distinct_keys must be positive.", std::invalid_argument);
Expand Down Expand Up @@ -170,7 +169,8 @@ void streaming_groupby::impl::initialize(table_view const& data, cuda::stream_re
decltype(_d_agg_results){raii.release(), +[](mutable_table_device_view* t) { t->destroy(); }};
}

_d_agg_kinds = cudf::detail::make_device_uvector_async(_agg_kinds, stream, mr);
_d_agg_kinds = std::make_unique<rmm::device_uvector<aggregation::Kind>>(
cudf::detail::make_device_uvector_async(_agg_kinds, stream, mr));

// Map each column in `values_view` back to its index in `data`.
_value_col_indices.reserve(values_view.num_columns());
Expand Down
2 changes: 1 addition & 1 deletion cpp/src/groupby/streaming_groupby/merge.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -138,7 +138,7 @@ void streaming_groupby::impl::do_merge(impl const& other, cuda::stream_ref strea
cuda::counting_iterator<int64_t>(0),
static_cast<int64_t>(other_distinct_keys) * num_agg_cols,
merge_single_pass_aggs_fn{
result.target_indices.begin(), _d_agg_kinds.data(), *d_source, *_d_agg_results});
result.target_indices.begin(), _d_agg_kinds->data(), *d_source, *_d_agg_results});
}

} // namespace cudf::groupby
Loading