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 | | // This file is adapted from |
19 | | // https://github.com/ClickHouse/ClickHouse/blob/master/src/AggregateFunctions/AggregateFunctionExponentialMovingAverage.cpp |
20 | | |
21 | | #pragma once |
22 | | |
23 | | #include <array> |
24 | | #include <cmath> |
25 | | #include <memory> |
26 | | |
27 | | #include "core/assert_cast.h" |
28 | | #include "core/block/column_with_type_and_name.h" |
29 | | #include "core/column/column_const.h" |
30 | | #include "core/column/column_nullable.h" |
31 | | #include "core/column/column_vector.h" |
32 | | #include "core/data_type/data_type_number.h" |
33 | | #include "core/types.h" |
34 | | #include "exprs/aggregate/aggregate_function.h" |
35 | | #include "exprs/vexpr_context.h" |
36 | | |
37 | | namespace doris { |
38 | | class Arena; |
39 | | class BufferReadable; |
40 | | class BufferWritable; |
41 | | class IColumn; |
42 | | |
43 | | /** |
44 | | * Exponentially smoothed moving average over time. |
45 | | * |
46 | | * Each value corresponds to a timeunit index. The half_decay parameter is the |
47 | | * time lag at which exponential weights decay by one-half. |
48 | | * |
49 | | * State is a (value, time) pair representing the exponentially accumulated sum |
50 | | * at a reference time. To get the average, divide by sumWeights(half_decay). |
51 | | * |
52 | | * Formula: |
53 | | * scale(dt, x) = 2^(-dt/x) |
54 | | * sumWeights(x) = 1 / (1 - 2^(-1/x)) |
55 | | * add(v, t): merge current state with point (v, t) |
56 | | * merge(a, b): move both to the later time, then sum values |
57 | | * get(): value / sumWeights(half_decay) |
58 | | * |
59 | | * Usage: exponential_moving_average(half_decay, value, timeunit) |
60 | | * - half_decay: constant double, the half-life period in timeunit units |
61 | | * - value: numeric column to average |
62 | | * - timeunit: numeric time index (not raw timestamp; use intDiv if needed) |
63 | | * Returns DOUBLE. |
64 | | */ |
65 | | struct ExponentialMovingAverageData { |
66 | | double value = 0.0; |
67 | | double time = 0.0; |
68 | | double half_decay = 0.0; |
69 | | |
70 | 0 | static double scale(double time_passed, double hd) { return std::exp2(-time_passed / hd); } |
71 | | |
72 | 0 | static double sum_weights(double hd) { return 1.0 / (1.0 - std::exp2(-1.0 / hd)); } |
73 | | |
74 | 0 | void add(double new_value, double current_time, double hd) { |
75 | 0 | half_decay = hd; |
76 | 0 | ExponentialMovingAverageData other; |
77 | 0 | other.value = new_value; |
78 | 0 | other.time = current_time; |
79 | 0 | merge_point(other, hd); |
80 | 0 | } |
81 | | |
82 | 0 | void merge_point(const ExponentialMovingAverageData& other, double hd) { |
83 | 0 | if (time > other.time) { |
84 | 0 | value = value + other.value * scale(time - other.time, hd); |
85 | 0 | } else if (time < other.time) { |
86 | 0 | value = other.value + value * scale(other.time - time, hd); |
87 | 0 | time = other.time; |
88 | 0 | } else { |
89 | 0 | value = value + other.value; |
90 | 0 | } |
91 | 0 | } |
92 | | |
93 | 0 | void merge(const ExponentialMovingAverageData& rhs) { |
94 | 0 | double hd = half_decay != 0.0 ? half_decay : rhs.half_decay; |
95 | 0 | if (hd == 0.0) { |
96 | 0 | return; |
97 | 0 | } |
98 | 0 | half_decay = hd; |
99 | 0 | merge_point(rhs, hd); |
100 | 0 | } |
101 | | |
102 | 0 | double get() const { |
103 | 0 | if (half_decay == 0.0) { |
104 | 0 | return 0.0; |
105 | 0 | } |
106 | 0 | return value / sum_weights(half_decay); |
107 | 0 | } |
108 | | |
109 | 0 | void write(BufferWritable& buf) const { |
110 | 0 | buf.write_binary(value); |
111 | 0 | buf.write_binary(time); |
112 | 0 | buf.write_binary(half_decay); |
113 | 0 | } |
114 | | |
115 | 0 | void read(BufferReadable& buf) { |
116 | 0 | buf.read_binary(value); |
117 | 0 | buf.read_binary(time); |
118 | 0 | buf.read_binary(half_decay); |
119 | 0 | } |
120 | | |
121 | 0 | void reset() { |
122 | 0 | value = 0.0; |
123 | 0 | time = 0.0; |
124 | 0 | half_decay = 0.0; |
125 | 0 | } |
126 | | }; |
127 | | |
128 | | class AggregateFunctionExponentialMovingAverage final |
129 | | : public IAggregateFunctionDataHelper<ExponentialMovingAverageData, |
130 | | AggregateFunctionExponentialMovingAverage>, |
131 | | MultiExpression, |
132 | | NullableAggregateFunction { |
133 | | public: |
134 | | AggregateFunctionExponentialMovingAverage(const DataTypes& argument_types_) |
135 | 0 | : IAggregateFunctionDataHelper<ExponentialMovingAverageData, |
136 | 0 | AggregateFunctionExponentialMovingAverage>( |
137 | 0 | argument_types_) {} |
138 | | |
139 | 0 | String get_name() const override { return "exponential_moving_average"; } |
140 | | |
141 | 0 | DataTypePtr get_return_type() const override { return std::make_shared<DataTypeFloat64>(); } |
142 | | |
143 | 0 | const std::vector<size_t>& get_const_argument_indexes() const override { |
144 | 0 | static const std::vector<size_t> indexes {0}; |
145 | 0 | return indexes; |
146 | 0 | } |
147 | | |
148 | 0 | void reset(AggregateDataPtr __restrict place) const override { this->data(place).reset(); } |
149 | | |
150 | | void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num, |
151 | 0 | Arena&) const override { |
152 | 0 | const double new_value = |
153 | 0 | assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[1]) |
154 | 0 | .get_data()[row_num]; |
155 | 0 | const double current_time = |
156 | 0 | assert_cast<const ColumnFloat64&, TypeCheckOnRelease::DISABLE>(*columns[2]) |
157 | 0 | .get_data()[row_num]; |
158 | 0 | const auto& half_decay_column = |
159 | 0 | *check_and_get_column_with_const<ColumnFloat64>(*columns[0]); |
160 | 0 | this->data(place).add(new_value, current_time, half_decay_column.get_data()[0]); |
161 | 0 | } |
162 | | |
163 | 0 | void check_input_columns_type(const IColumn** columns) const override { |
164 | 0 | this->template check_const_argument_column_type<ColumnFloat64>(columns[0]); |
165 | 0 | this->template check_argument_column_type<ColumnFloat64>(columns[1]); |
166 | 0 | this->template check_argument_column_type<ColumnFloat64>(columns[2]); |
167 | 0 | } |
168 | | |
169 | | void merge(AggregateDataPtr __restrict place, ConstAggregateDataPtr rhs, |
170 | 0 | Arena&) const override { |
171 | 0 | this->data(place).merge(this->data(rhs)); |
172 | 0 | } |
173 | | |
174 | 0 | void serialize(ConstAggregateDataPtr __restrict place, BufferWritable& buf) const override { |
175 | 0 | this->data(place).write(buf); |
176 | 0 | } |
177 | | |
178 | | void deserialize(AggregateDataPtr __restrict place, BufferReadable& buf, |
179 | 0 | Arena&) const override { |
180 | 0 | this->data(place).read(buf); |
181 | 0 | } |
182 | | |
183 | 0 | void insert_result_into(ConstAggregateDataPtr __restrict place, IColumn& to) const override { |
184 | 0 | assert_cast<ColumnFloat64&, TypeCheckOnRelease::DISABLE>(to).get_data().push_back( |
185 | 0 | this->data(place).get()); |
186 | 0 | } |
187 | | }; |
188 | | |
189 | | } // namespace doris |