arrow/acero/aggregate_node.h

0001 // Licensed to the Apache Software Foundation (ASF) under one
0002 // or more contributor license agreements.  See the NOTICE file
0003 // distributed with this work for additional information
0004 // regarding copyright ownership.  The ASF licenses this file
0005 // to you under the Apache License, Version 2.0 (the
0006 // "License"); you may not use this file except in compliance
0007 // with the License.  You may obtain a copy of the License at
0008 //
0009 //   http://www.apache.org/licenses/LICENSE-2.0
0010 //
0011 // Unless required by applicable law or agreed to in writing,
0012 // software distributed under the License is distributed on an
0013 // "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
0014 // KIND, either express or implied.  See the License for the
0015 // specific language governing permissions and limitations
0016 // under the License.
0017
0018 // This API is EXPERIMENTAL.
0019
0020 #pragma once
0021
0022 #include <memory>
0023 #include <vector>
0024
0025 #include "arrow/acero/visibility.h"
0026 #include "arrow/compute/api_aggregate.h"
0027 #include "arrow/compute/type_fwd.h"
0028 #include "arrow/result.h"
0029 #include "arrow/type_fwd.h"
0030
0031 namespace arrow {
0032 namespace acero {
0033 namespace aggregate {
0034
0035 using compute::Aggregate;
0036 using compute::default_exec_context;
0037 using compute::ExecContext;
0038
0039 /// \brief Make the output schema of an aggregate node
0040 ///
0041 /// The output schema is determined by the aggregation kernels, which may depend on the
0042 /// ExecContext argument. To guarantee correct results, the same ExecContext argument
0043 /// should be used in execution.
0044 ///
0045 /// \param[in] input_schema the schema of the input to the node
0046 /// \param[in] keys the grouping keys for the aggregation
0047 /// \param[in] segment_keys the segmenting keys for the aggregation
0048 /// \param[in] aggregates the aggregates for the aggregation
0049 /// \param[in] exec_ctx the execution context for the aggregation
0050 ARROW_ACERO_EXPORT Result<std::shared_ptr<Schema>> MakeOutputSchema(
0051     const std::shared_ptr<Schema>& input_schema, const std::vector<FieldRef>& keys,
0052     const std::vector<FieldRef>& segment_keys, const std::vector<Aggregate>& aggregates,
0053     ExecContext* exec_ctx = default_exec_context());
0054
0055 }  // namespace aggregate
0056 }  // namespace acero
0057 }  // namespace arrow