be/src/exprs/aggregate/aggregate_function_percentile_reservoir.h
Line | Count | Source |
1 | | // Licensed to the Apache Software Foundation (ASF) under one |
2 | | // or more contributor license agreements. See the NOTICE file |
3 | | // distributed with this work for additional information |
4 | | // regarding copyright ownership. The ASF licenses this file |
5 | | // to you under the Apache License, Version 2.0 (the |
6 | | // "License"); you may not use this file except in compliance |
7 | | // with the License. You may obtain a copy of the License at |
8 | | // |
9 | | // http://www.apache.org/licenses/LICENSE-2.0 |
10 | | // |
11 | | // Unless required by applicable law or agreed to in writing, |
12 | | // software distributed under the License is distributed on an |
13 | | // "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
14 | | // KIND, either express or implied. See the License for the |
15 | | // specific language governing permissions and limitations |
16 | | // under the License. |
17 | | |
18 | | #pragma once |
19 | | |
20 | | #include <glog/logging.h> |
21 | | |
22 | | #include <memory> |
23 | | |
24 | | #include "core/column/column_vector.h" |
25 | | #include "core/data_type/data_type_number.h" |
26 | | #include "core/data_type/primitive_type.h" |
27 | | #include "exprs/aggregate/aggregate_function.h" |
28 | | #include "util/reservoir_sampler.h" |
29 | | |
30 | | namespace doris { |
31 | | |
32 | | class Arena; |
33 | | class BufferReadable; |
34 | | |
35 | | struct QuantileReservoirSampler { |
36 | 0 | void add(const double x, const double input_level) { |
37 | 0 | this->level = input_level; |
38 | 0 | data.insert(x); |
39 | 0 | } |
40 | | |
41 | 2 | void add_batch(const double* values, size_t size, const double input_level) { |
42 | 2 | this->level = input_level; |
43 | 2 | data.insert_many(values, size); |
44 | 2 | } |
45 | | |
46 | 0 | void merge(const QuantileReservoirSampler& rhs) { |
47 | 0 | level = rhs.level; |
48 | 0 | data.merge(rhs.data); |
49 | 0 | } |
50 | | |
51 | 2 | void reset() { |
52 | 2 | level = 0.0; |
53 | 2 | data.clear(); |
54 | 2 | } |
55 | | |
56 | 0 | void serialize(BufferWritable& buf) const { |
57 | 0 | buf.write_binary(level); |
58 | 0 | data.write(buf); |
59 | 0 | } |
60 | | |
61 | 0 | void deserialize(BufferReadable& buf) { |
62 | 0 | buf.read_binary(level); |
63 | 0 | data.read(buf); |
64 | 0 | } |
65 | | |
66 | 2 | double get() const { |
67 | | // The caller is a ConstAggregateDataPtr, but it itself is an AggregateDataPtr. |
68 | | // To call a non-const method here, a const_cast is required. |
69 | 2 | return const_cast<ReservoirSampler&>(data).quantileInterpolated(this->level); |
70 | 2 | } |
71 | | |
72 | | private: |
73 | | double level = 0.0; |
74 | | ReservoirSampler data; |
75 | | }; |
76 | | |
77 | | template <typename Data> |
78 | | class AggregateFunctionPercentileReservoir final |
79 | | : public IAggregateFunctionDataHelper<Data, AggregateFunctionPercentileReservoir<Data>>, |
80 | | MultiExpression, |
81 | | NullableAggregateFunction { |
82 | | public: |
83 | | AggregateFunctionPercentileReservoir(const DataTypes& argument_types_) |
84 | 1 | : IAggregateFunctionDataHelper<Data, AggregateFunctionPercentileReservoir<Data>>( |
85 | 1 | argument_types_) {} |
86 | | |
87 | 0 | String get_name() const override { return "percentile_reservoir"; } |
88 | | |
89 | 0 | DataTypePtr get_return_type() const override { return std::make_shared<DataTypeFloat64>(); } |
90 | | |
91 | | void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, |
92 | 0 | Arena&) const override { |
93 | 0 | auto value = assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[0]) |
94 | 0 | .get_data()[row_num]; |
95 | 0 | auto level = assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[1]) |
96 | 0 | .get_data()[0]; |
97 | 0 | this->data(place).add(value, level); |
98 | 0 | } |
99 | | |
100 | 0 | void check_input_columns_type(const IColumn** columns) const override { |
101 | 0 | this->template check_argument_column_type<ColumnFloat64>(columns[0]); |
102 | 0 | this->template check_argument_column_type<ColumnFloat64>(columns[1]); |
103 | 0 | } |
104 | | |
105 | | void add_batch_single_place(size_t batch_size, AggregateDataPtr place, const IColumn** columns, |
106 | 1 | Arena&) const override { |
107 | 1 | const auto& sources = |
108 | 1 | assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[0]); |
109 | 1 | const auto& levels = |
110 | 1 | assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[1]); |
111 | 1 | this->data(place).add_batch(sources.get_data().data(), batch_size, levels.get_data()[0]); |
112 | 1 | } |
113 | | |
114 | | void add_range_single_place(int64_t partition_start, int64_t partition_end, int64_t frame_start, |
115 | | int64_t frame_end, AggregateDataPtr place, const IColumn** columns, |
116 | | Arena&, UInt8* use_null_result, |
117 | 2 | UInt8* could_use_previous_result) const override { |
118 | 2 | frame_start = std::max<int64_t>(frame_start, partition_start); |
119 | 2 | frame_end = std::min<int64_t>(frame_end, partition_end); |
120 | 2 | if (frame_start < frame_end) { |
121 | 1 | const auto& sources = |
122 | 1 | assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[0]); |
123 | 1 | const auto& levels = |
124 | 1 | assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[1]); |
125 | 1 | this->data(place).add_batch(sources.get_data().data() + frame_start, |
126 | 1 | frame_end - frame_start, levels.get_data()[0]); |
127 | 1 | *use_null_result = false; |
128 | 1 | *could_use_previous_result = true; |
129 | 1 | } else if (!*could_use_previous_result) { |
130 | 1 | *use_null_result = true; |
131 | 1 | } |
132 | 2 | } |
133 | | |
134 | 2 | void reset(AggregateDataPtr place) const override { this->data(place).reset(); } |
135 | | |
136 | | void merge(AggregateDataPtr __restrict place, ConstAggregateDataPtr rhs, |
137 | 0 | Arena&) const override { |
138 | 0 | this->data(place).merge(this->data(rhs)); |
139 | 0 | } |
140 | | |
141 | 0 | void serialize(ConstAggregateDataPtr __restrict place, BufferWritable& buf) const override { |
142 | 0 | this->data(place).serialize(buf); |
143 | 0 | } |
144 | | |
145 | | void deserialize(AggregateDataPtr __restrict place, BufferReadable& buf, |
146 | 0 | Arena&) const override { |
147 | 0 | this->data(place).deserialize(buf); |
148 | 0 | } |
149 | | |
150 | 2 | void insert_result_into(ConstAggregateDataPtr __restrict place, IColumn& to) const override { |
151 | 2 | assert_cast<ColumnFloat64&, TypeCheckOnRelease::DISABLE>(to).get_data().push_back( |
152 | 2 | this->data(place).get()); |
153 | 2 | } |
154 | | }; |
155 | | |
156 | | } // namespace doris |