Coverage Report

Created: 2026-03-12 17:15

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
be/src/exprs/aggregate/aggregate_function_stddev.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 <boost/iterator/iterator_facade.hpp>
21
#include <cmath>
22
#include <cstddef>
23
#include <cstdint>
24
#include <memory>
25
#include <type_traits>
26
27
#include "core/assert_cast.h"
28
#include "core/column/column.h"
29
#include "core/column/column_nullable.h"
30
#include "core/data_type/data_type_decimal.h"
31
#include "core/data_type/data_type_number.h"
32
#include "core/types.h"
33
#include "exprs/aggregate/aggregate_function.h"
34
35
namespace doris {
36
#include "common/compile_check_begin.h"
37
class Arena;
38
class BufferReadable;
39
class BufferWritable;
40
template <PrimitiveType T>
41
class ColumnDecimal;
42
template <PrimitiveType T>
43
class ColumnVector;
44
45
template <PrimitiveType T, bool is_stddev>
46
struct BaseData {
47
0
    BaseData() = default;
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EEC2Ev
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EEC2Ev
48
0
    virtual ~BaseData() = default;
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EED2Ev
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EED2Ev
49
50
0
    void write(BufferWritable& buf) const {
51
0
        buf.write_binary(mean);
52
0
        buf.write_binary(m2);
53
0
        buf.write_binary(count);
54
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE5writeERNS_14BufferWritableE
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE5writeERNS_14BufferWritableE
55
56
0
    void read(BufferReadable& buf) {
57
0
        buf.read_binary(mean);
58
0
        buf.read_binary(m2);
59
0
        buf.read_binary(count);
60
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE4readERNS_14BufferReadableE
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE4readERNS_14BufferReadableE
61
62
0
    void reset() {
63
0
        mean = 0.0;
64
0
        m2 = 0.0;
65
0
        count = 0;
66
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE5resetEv
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE5resetEv
67
68
0
    double get_result(double res) const {
69
0
        auto inf_to_nan = [](double val) {
70
            // This function performs squaring operations, and due to differences in computation order,
71
            // it might produce different values such as inf and nan.
72
            // In MySQL, this will directly result in an error due to exceeding the double range.
73
            // For performance reasons, we are uniformly changing it to nan
74
0
            if (std::isinf(val)) {
75
0
                return std::nan("");
76
0
            }
77
0
            return val;
78
0
        };
Unexecuted instantiation: _ZZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE10get_resultEdENKUldE_clEd
Unexecuted instantiation: _ZZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE10get_resultEdENKUldE_clEd
79
0
        if constexpr (is_stddev) {
80
0
            return inf_to_nan(std::sqrt(res));
81
0
        } else {
82
0
            return inf_to_nan(res);
83
0
        }
84
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE10get_resultEd
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE10get_resultEd
85
86
0
    double get_pop_result() const {
87
0
        if (count == 1) {
88
0
            return 0.0;
89
0
        }
90
0
        double res = m2 / (double)count;
91
0
        return get_result(res);
92
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE14get_pop_resultEv
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE14get_pop_resultEv
93
94
0
    double get_samp_result() const {
95
0
        double res = m2 / double(count - 1);
96
0
        return get_result(res);
97
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE15get_samp_resultEv
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE15get_samp_resultEv
98
99
0
    void merge(const BaseData& rhs) {
100
0
        if (rhs.count == 0) {
101
0
            return;
102
0
        }
103
0
        double delta = mean - rhs.mean;
104
0
        double sum_count = double(count + rhs.count);
105
0
        mean = rhs.mean + delta * (double)count / sum_count;
106
0
        m2 = rhs.m2 + m2 + (delta * delta) * (double)rhs.count * (double)count / sum_count;
107
0
        count = int64_t(sum_count);
108
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE5mergeERKS2_
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE5mergeERKS2_
109
110
0
    void add(const IColumn* column, size_t row_num) {
111
0
        const auto& sources = assert_cast<const typename PrimitiveTypeTraits<T>::ColumnType&,
112
0
                                          TypeCheckOnRelease::DISABLE>(*column);
113
0
        double source_data = (double)sources.get_data()[row_num];
114
115
0
        double delta = source_data - mean;
116
0
        double r = delta / double(1 + count);
117
0
        mean += r;
118
0
        m2 += (double)count * delta * r;
119
0
        count += 1;
120
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE3addEPKNS_7IColumnEm
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE3addEPKNS_7IColumnEm
121
122
    double mean {};
123
    double m2 {};
124
    int64_t count {};
125
};
126
127
template <PrimitiveType T, typename Name, bool is_stddev>
128
struct PopData : BaseData<T, is_stddev>, Name {
129
    using ColVecResult = std::conditional_t<is_decimal(T), ColumnDecimal128V2, ColumnFloat64>;
130
0
    void insert_result_into(IColumn& to) const {
131
0
        auto& col = assert_cast<ColVecResult&>(to);
132
        if constexpr (is_decimal(T)) {
133
            col.get_data().push_back(this->get_pop_result().value());
134
0
        } else {
135
0
            col.get_data().push_back(this->get_pop_result());
136
0
        }
137
0
    }
Unexecuted instantiation: _ZNK5doris7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EE18insert_result_intoERNS_7IColumnE
Unexecuted instantiation: _ZNK5doris7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EE18insert_result_intoERNS_7IColumnE
138
139
0
    static DataTypePtr get_return_type() { return std::make_shared<DataTypeFloat64>(); }
Unexecuted instantiation: _ZN5doris7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EE15get_return_typeEv
Unexecuted instantiation: _ZN5doris7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EE15get_return_typeEv
140
};
141
142
// For this series of functions, the Decimal type is not supported
143
// because the operations involve squaring,
144
// which can easily exceed the range of the Decimal type.
145
146
template <PrimitiveType T, typename Name, bool is_stddev>
147
struct SampData : BaseData<T, is_stddev>, Name {
148
    using ColVecResult = std::conditional_t<is_decimal(T), ColumnDecimal128V2, ColumnFloat64>;
149
0
    void insert_result_into(IColumn& to) const {
150
0
        auto& col = assert_cast<ColVecResult&>(to);
151
0
        if (this->count == 1 || this->count == 0) {
152
0
            col.insert_default();
153
0
        } else {
154
            if constexpr (is_decimal(T)) {
155
                col.get_data().push_back(this->get_samp_result().value());
156
0
            } else {
157
0
                col.get_data().push_back(this->get_samp_result());
158
0
            }
159
0
        }
160
0
    }
Unexecuted instantiation: _ZNK5doris8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EE18insert_result_intoERNS_7IColumnE
Unexecuted instantiation: _ZNK5doris8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EE18insert_result_intoERNS_7IColumnE
161
162
0
    static DataTypePtr get_return_type() { return std::make_shared<DataTypeFloat64>(); }
Unexecuted instantiation: _ZN5doris8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EE15get_return_typeEv
Unexecuted instantiation: _ZN5doris8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EE15get_return_typeEv
163
};
164
165
struct StddevName {
166
0
    static const char* name() { return "stddev"; }
167
};
168
struct VarianceName {
169
0
    static const char* name() { return "variance"; }
170
};
171
struct VarianceSampName {
172
0
    static const char* name() { return "variance_samp"; }
173
};
174
struct StddevSampName {
175
0
    static const char* name() { return "stddev_samp"; }
176
};
177
178
template <typename Data>
179
class AggregateFunctionSampVariance
180
        : public IAggregateFunctionDataHelper<Data, AggregateFunctionSampVariance<Data>>,
181
          UnaryExpression,
182
          NullableAggregateFunction {
183
public:
184
    AggregateFunctionSampVariance(const DataTypes& argument_types_)
185
0
            : IAggregateFunctionDataHelper<Data, AggregateFunctionSampVariance<Data>>(
186
0
                      argument_types_) {}
Unexecuted instantiation: _ZN5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEEC2ERKSt6vectorISt10shared_ptrIKNS_9IDataTypeEESaISA_EE
Unexecuted instantiation: _ZN5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEEC2ERKSt6vectorISt10shared_ptrIKNS_9IDataTypeEESaISA_EE
Unexecuted instantiation: _ZN5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEEC2ERKSt6vectorISt10shared_ptrIKNS_9IDataTypeEESaISA_EE
Unexecuted instantiation: _ZN5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEEC2ERKSt6vectorISt10shared_ptrIKNS_9IDataTypeEESaISA_EE
187
188
0
    String get_name() const override { return Data::name(); }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE8get_nameB5cxx11Ev
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE8get_nameB5cxx11Ev
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE8get_nameB5cxx11Ev
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE8get_nameB5cxx11Ev
189
190
0
    DataTypePtr get_return_type() const override { return Data::get_return_type(); }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE15get_return_typeEv
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE15get_return_typeEv
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE15get_return_typeEv
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE15get_return_typeEv
191
192
    void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num,
193
0
             Arena&) const override {
194
0
        this->data(place).add(columns[0], row_num);
195
0
    }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE3addEPcPPKNS_7IColumnElRNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE3addEPcPPKNS_7IColumnElRNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE3addEPcPPKNS_7IColumnElRNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE3addEPcPPKNS_7IColumnElRNS_5ArenaE
196
197
0
    void reset(AggregateDataPtr __restrict place) const override { this->data(place).reset(); }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE5resetEPc
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE5resetEPc
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE5resetEPc
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE5resetEPc
198
199
    void merge(AggregateDataPtr __restrict place, ConstAggregateDataPtr rhs,
200
0
               Arena&) const override {
201
0
        this->data(place).merge(this->data(rhs));
202
0
    }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE5mergeEPcPKcRNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE5mergeEPcPKcRNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE5mergeEPcPKcRNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE5mergeEPcPKcRNS_5ArenaE
203
204
0
    void serialize(ConstAggregateDataPtr __restrict place, BufferWritable& buf) const override {
205
0
        this->data(place).write(buf);
206
0
    }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE9serializeEPKcRNS_14BufferWritableE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE9serializeEPKcRNS_14BufferWritableE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE9serializeEPKcRNS_14BufferWritableE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE9serializeEPKcRNS_14BufferWritableE
207
208
    void deserialize(AggregateDataPtr __restrict place, BufferReadable& buf,
209
0
                     Arena&) const override {
210
0
        this->data(place).read(buf);
211
0
    }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE11deserializeEPcRNS_14BufferReadableERNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE11deserializeEPcRNS_14BufferReadableERNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE11deserializeEPcRNS_14BufferReadableERNS_5ArenaE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE11deserializeEPcRNS_14BufferReadableERNS_5ArenaE
212
213
0
    void insert_result_into(ConstAggregateDataPtr __restrict place, IColumn& to) const override {
214
0
        this->data(place).insert_result_into(to);
215
0
    }
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EEEE18insert_result_intoEPKcRNS_7IColumnE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EEEE18insert_result_intoEPKcRNS_7IColumnE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EEEE18insert_result_intoEPKcRNS_7IColumnE
Unexecuted instantiation: _ZNK5doris29AggregateFunctionSampVarianceINS_8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EEEE18insert_result_intoEPKcRNS_7IColumnE
216
};
217
218
} // namespace doris
219
220
#include "common/compile_check_end.h"