1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
|
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.spark.sql
import scala.language.postfixOps
import org.apache.spark.sql.expressions.Aggregator
import org.apache.spark.sql.functions._
import org.apache.spark.sql.test.SharedSQLContext
/** An `Aggregator` that adds up any numeric type returned by the given function. */
class SumOf[I, N : Numeric](f: I => N) extends Aggregator[I, N, N] {
val numeric = implicitly[Numeric[N]]
override def zero: N = numeric.zero
override def reduce(b: N, a: I): N = numeric.plus(b, f(a))
override def merge(b1: N, b2: N): N = numeric.plus(b1, b2)
override def finish(reduction: N): N = reduction
}
object TypedAverage extends Aggregator[(String, Int), (Long, Long), Double] {
override def zero: (Long, Long) = (0, 0)
override def reduce(countAndSum: (Long, Long), input: (String, Int)): (Long, Long) = {
(countAndSum._1 + 1, countAndSum._2 + input._2)
}
override def merge(b1: (Long, Long), b2: (Long, Long)): (Long, Long) = {
(b1._1 + b2._1, b1._2 + b2._2)
}
override def finish(countAndSum: (Long, Long)): Double = countAndSum._2 / countAndSum._1
}
object ComplexResultAgg extends Aggregator[(String, Int), (Long, Long), (Long, Long)] {
override def zero: (Long, Long) = (0, 0)
override def reduce(countAndSum: (Long, Long), input: (String, Int)): (Long, Long) = {
(countAndSum._1 + 1, countAndSum._2 + input._2)
}
override def merge(b1: (Long, Long), b2: (Long, Long)): (Long, Long) = {
(b1._1 + b2._1, b1._2 + b2._2)
}
override def finish(reduction: (Long, Long)): (Long, Long) = reduction
}
case class AggData(a: Int, b: String)
object ClassInputAgg extends Aggregator[AggData, Int, Int] {
/** A zero value for this aggregation. Should satisfy the property that any b + zero = b */
override def zero: Int = 0
/**
* Combine two values to produce a new value. For performance, the function may modify `b` and
* return it instead of constructing new object for b.
*/
override def reduce(b: Int, a: AggData): Int = b + a.a
/**
* Transform the output of the reduction.
*/
override def finish(reduction: Int): Int = reduction
/**
* Merge two intermediate values
*/
override def merge(b1: Int, b2: Int): Int = b1 + b2
}
object ComplexBufferAgg extends Aggregator[AggData, (Int, AggData), Int] {
/** A zero value for this aggregation. Should satisfy the property that any b + zero = b */
override def zero: (Int, AggData) = 0 -> AggData(0, "0")
/**
* Combine two values to produce a new value. For performance, the function may modify `b` and
* return it instead of constructing new object for b.
*/
override def reduce(b: (Int, AggData), a: AggData): (Int, AggData) = (b._1 + 1, a)
/**
* Transform the output of the reduction.
*/
override def finish(reduction: (Int, AggData)): Int = reduction._1
/**
* Merge two intermediate values
*/
override def merge(b1: (Int, AggData), b2: (Int, AggData)): (Int, AggData) =
(b1._1 + b2._1, b1._2)
}
class DatasetAggregatorSuite extends QueryTest with SharedSQLContext {
import testImplicits._
def sum[I, N : Numeric : Encoder](f: I => N): TypedColumn[I, N] =
new SumOf(f).toColumn
test("typed aggregation: TypedAggregator") {
val ds = Seq(("a", 10), ("a", 20), ("b", 1), ("b", 2), ("c", 1)).toDS()
checkDataset(
ds.groupByKey(_._1).agg(sum(_._2)),
("a", 30), ("b", 3), ("c", 1))
}
test("typed aggregation: TypedAggregator, expr, expr") {
val ds = Seq(("a", 10), ("a", 20), ("b", 1), ("b", 2), ("c", 1)).toDS()
checkDataset(
ds.groupByKey(_._1).agg(
sum(_._2),
expr("sum(_2)").as[Long],
count("*")),
("a", 30, 30L, 2L), ("b", 3, 3L, 2L), ("c", 1, 1L, 1L))
}
test("typed aggregation: complex case") {
val ds = Seq("a" -> 1, "a" -> 3, "b" -> 3).toDS()
checkDataset(
ds.groupByKey(_._1).agg(
expr("avg(_2)").as[Double],
TypedAverage.toColumn),
("a", 2.0, 2.0), ("b", 3.0, 3.0))
}
test("typed aggregation: complex result type") {
val ds = Seq("a" -> 1, "a" -> 3, "b" -> 3).toDS()
checkDataset(
ds.groupByKey(_._1).agg(
expr("avg(_2)").as[Double],
ComplexResultAgg.toColumn),
("a", 2.0, (2L, 4L)), ("b", 3.0, (1L, 3L)))
}
test("typed aggregation: in project list") {
val ds = Seq(1, 3, 2, 5).toDS()
checkDataset(
ds.select(sum((i: Int) => i)),
11)
checkDataset(
ds.select(sum((i: Int) => i), sum((i: Int) => i * 2)),
11 -> 22)
}
test("typed aggregation: class input") {
val ds = Seq(AggData(1, "one"), AggData(2, "two")).toDS()
checkDataset(
ds.select(ClassInputAgg.toColumn),
3)
}
test("typed aggregation: class input with reordering") {
val ds = sql("SELECT 'one' AS b, 1 as a").as[AggData]
checkDataset(
ds.select(ClassInputAgg.toColumn),
1)
checkDataset(
ds.select(expr("avg(a)").as[Double], ClassInputAgg.toColumn),
(1.0, 1))
checkDataset(
ds.groupByKey(_.b).agg(ClassInputAgg.toColumn),
("one", 1))
}
test("typed aggregation: complex input") {
val ds = Seq(AggData(1, "one"), AggData(2, "two")).toDS()
checkDataset(
ds.select(ComplexBufferAgg.toColumn),
2
)
checkDataset(
ds.select(expr("avg(a)").as[Double], ComplexBufferAgg.toColumn),
(1.5, 2))
checkDataset(
ds.groupByKey(_.b).agg(ComplexBufferAgg.toColumn),
("one", 1), ("two", 1))
}
}
|
