pitrou commented on code in PR #45562:
URL: https://github.com/apache/arrow/pull/45562#discussion_r1977539472
##########
cpp/src/arrow/compute/kernels/hash_aggregate.cc:
##########
@@ -3319,9 +3324,401 @@ struct GroupedListFactory {
HashAggregateKernel kernel;
InputType argument_type;
};
-} // namespace
-namespace {
+// ----------------------------------------------------------------------
+// Pivot implementation
+
+struct GroupedPivotAccumulator {
+ Status Init(ExecContext* ctx, std::shared_ptr<DataType> value_type,
+ const PivotWiderOptions* options) {
+ ctx_ = ctx;
+ value_type_ = std::move(value_type);
+ num_keys_ = static_cast<int>(options->key_names.size());
+ num_groups_ = 0;
+ columns_.resize(num_keys_);
+ scratch_buffer_ = BufferBuilder(ctx_->memory_pool());
+ return Status::OK();
+ }
+
+ Status Consume(span<const uint32_t> groups, span<const PivotWiderKeyIndex>
keys,
+ const ArraySpan& values) {
+ // To dispatch the values into the right (group, key) coordinates,
+ // we first compute a vector of take indices for each output column.
+ //
+ // For each index #i, we set take_indices[keys[#i]][groups[#i]] = #i.
+ // Unpopulated take_indices entries are null.
+ //
+ // For example, assuming we get:
+ // groups | keys
+ // ===================
+ // 1 | 0
+ // 3 | 1
+ // 1 | 1
+ // 0 | 1
+ //
+ // We are going to compute:
+ // - take_indices[key = 0] = [null, 0, null, null]
+ // - take_indices[key = 1] = [3, 2, null, 2]
Review Comment:
Nice catch!
--
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]
