Coverage Report

Created: 2026-05-29 19:04

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 <limits>
25
#include <memory>
26
27
#include "core/assert_cast.h"
28
#include "core/column/column.h"
29
#include "core/data_type/data_type_number.h"
30
#include "core/types.h"
31
#include "exprs/aggregate/aggregate_function.h"
32
33
namespace doris {
34
class Arena;
35
class BufferReadable;
36
class BufferWritable;
37
template <PrimitiveType T>
38
class ColumnVector;
39
40
template <PrimitiveType T, bool is_stddev>
41
struct BaseData {
42
0
    BaseData() = default;
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EEC2Ev
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EEC2Ev
43
0
    virtual ~BaseData() = default;
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EED2Ev
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EED2Ev
44
45
0
    void write(BufferWritable& buf) const {
46
0
        buf.write_binary(mean);
47
0
        buf.write_binary(m2);
48
0
        buf.write_binary(count);
49
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE5writeERNS_14BufferWritableE
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE5writeERNS_14BufferWritableE
50
51
0
    void read(BufferReadable& buf) {
52
0
        buf.read_binary(mean);
53
0
        buf.read_binary(m2);
54
0
        buf.read_binary(count);
55
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE4readERNS_14BufferReadableE
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE4readERNS_14BufferReadableE
56
57
0
    void reset() {
58
0
        mean = 0.0;
59
0
        m2 = 0.0;
60
0
        count = 0;
61
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE5resetEv
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE5resetEv
62
63
0
    double get_result(double res) const {
64
0
        auto inf_to_nan = [](double val) {
65
            // This function performs squaring operations, and due to differences in computation order,
66
            // it might produce different values such as inf and nan.
67
            // In MySQL, this will directly result in an error due to exceeding the double range.
68
            // For performance reasons, we are uniformly changing it to nan
69
0
            if (std::isinf(val)) {
70
0
                return std::numeric_limits<double>::quiet_NaN();
71
0
            }
72
0
            return val;
73
0
        };
Unexecuted instantiation: _ZZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE10get_resultEdENKUldE_clEd
Unexecuted instantiation: _ZZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE10get_resultEdENKUldE_clEd
74
0
        if constexpr (is_stddev) {
75
0
            return inf_to_nan(std::sqrt(res));
76
0
        } else {
77
0
            return inf_to_nan(res);
78
0
        }
79
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE10get_resultEd
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE10get_resultEd
80
81
0
    double get_pop_result() const {
82
0
        if (count == 1) {
83
0
            return 0.0;
84
0
        }
85
0
        double res = m2 / (double)count;
86
0
        return get_result(res);
87
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE14get_pop_resultEv
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE14get_pop_resultEv
88
89
0
    double get_samp_result() const {
90
0
        double res = m2 / double(count - 1);
91
0
        return get_result(res);
92
0
    }
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE15get_samp_resultEv
Unexecuted instantiation: _ZNK5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE15get_samp_resultEv
93
94
0
    void merge(const BaseData& rhs) {
95
0
        if (rhs.count == 0) {
96
0
            return;
97
0
        }
98
0
        double delta = mean - rhs.mean;
99
0
        double sum_count = double(count + rhs.count);
100
0
        mean = rhs.mean + delta * (double)count / sum_count;
101
0
        m2 = rhs.m2 + m2 + (delta * delta) * (double)rhs.count * (double)count / sum_count;
102
0
        count = int64_t(sum_count);
103
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE5mergeERKS2_
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE5mergeERKS2_
104
105
0
    void add(const IColumn* column, size_t row_num) {
106
0
        const auto& sources = assert_cast<const typename PrimitiveTypeTraits<T>::ColumnType&,
107
0
                                          TypeCheckOnRelease::DISABLE>(*column);
108
0
        double source_data = (double)sources.get_data()[row_num];
109
110
0
        double delta = source_data - mean;
111
0
        double r = delta / double(1 + count);
112
0
        mean += r;
113
0
        m2 += (double)count * delta * r;
114
0
        count += 1;
115
0
    }
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb0EE3addEPKNS_7IColumnEm
Unexecuted instantiation: _ZN5doris8BaseDataILNS_13PrimitiveTypeE9ELb1EE3addEPKNS_7IColumnEm
116
117
    double mean {};
118
    double m2 {};
119
    int64_t count {};
120
};
121
122
template <PrimitiveType T, typename Name, bool is_stddev>
123
struct PopData : BaseData<T, is_stddev>, Name {
124
0
    void insert_result_into(IColumn& to) const {
125
0
        auto& col = assert_cast<ColumnFloat64&>(to);
126
0
        col.get_data().push_back(this->get_pop_result());
127
0
    }
Unexecuted instantiation: _ZNK5doris7PopDataILNS_13PrimitiveTypeE9ENS_12VarianceNameELb0EE18insert_result_intoERNS_7IColumnE
Unexecuted instantiation: _ZNK5doris7PopDataILNS_13PrimitiveTypeE9ENS_10StddevNameELb1EE18insert_result_intoERNS_7IColumnE
128
129
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
130
};
131
132
// For this series of functions, the Decimal type is not supported
133
// because the operations involve squaring,
134
// which can easily exceed the range of the Decimal type.
135
136
template <PrimitiveType T, typename Name, bool is_stddev>
137
struct SampData : BaseData<T, is_stddev>, Name {
138
0
    void insert_result_into(IColumn& to) const {
139
0
        auto& col = assert_cast<ColumnFloat64&>(to);
140
0
        if (this->count == 1 || this->count == 0) {
141
0
            col.get_data().push_back(std::numeric_limits<double>::quiet_NaN());
142
0
        } else {
143
0
            col.get_data().push_back(this->get_samp_result());
144
0
        }
145
0
    }
Unexecuted instantiation: _ZNK5doris8SampDataILNS_13PrimitiveTypeE9ENS_16VarianceSampNameELb0EE18insert_result_intoERNS_7IColumnE
Unexecuted instantiation: _ZNK5doris8SampDataILNS_13PrimitiveTypeE9ENS_14StddevSampNameELb1EE18insert_result_intoERNS_7IColumnE
146
147
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
148
};
149
150
struct StddevName {
151
0
    static const char* name() { return "stddev"; }
152
};
153
struct VarianceName {
154
0
    static const char* name() { return "variance"; }
155
};
156
struct VarianceSampName {
157
0
    static const char* name() { return "variance_samp"; }
158
};
159
struct StddevSampName {
160
0
    static const char* name() { return "stddev_samp"; }
161
};
162
163
template <typename Data>
164
class AggregateFunctionSampVariance final
165
        : public IAggregateFunctionDataHelper<Data, AggregateFunctionSampVariance<Data>>,
166
          UnaryExpression,
167
          NullableAggregateFunction {
168
public:
169
    AggregateFunctionSampVariance(const DataTypes& argument_types_)
170
0
            : IAggregateFunctionDataHelper<Data, AggregateFunctionSampVariance<Data>>(
171
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
172
173
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
174
175
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
176
177
    void add(AggregateDataPtr __restrict place, const IColumn** columns, ssize_t row_num,
178
0
             Arena&) const override {
179
0
        this->data(place).add(columns[0], row_num);
180
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
181
182
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
183
184
    void merge(AggregateDataPtr __restrict place, ConstAggregateDataPtr rhs,
185
0
               Arena&) const override {
186
0
        this->data(place).merge(this->data(rhs));
187
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
188
189
0
    void serialize(ConstAggregateDataPtr __restrict place, BufferWritable& buf) const override {
190
0
        this->data(place).write(buf);
191
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
192
193
    void deserialize(AggregateDataPtr __restrict place, BufferReadable& buf,
194
0
                     Arena&) const override {
195
0
        this->data(place).read(buf);
196
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
197
198
0
    void insert_result_into(ConstAggregateDataPtr __restrict place, IColumn& to) const override {
199
0
        this->data(place).insert_result_into(to);
200
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
201
};
202
203
} // namespace doris