Funciones de agregado definidas por el usuario (UDAF)

Se aplica a: Databricks Runtime

Las funciones de agregado definidas por el usuario (UDAF) son rutinas programables por el usuario que actúan en varias filas a la vez y devuelven un único valor agregado como resultado. En esta documentación se enumeran las clases necesarias para crear y registrar UDAF. También contiene ejemplos que muestran cómo definir y registrar UDAF en Scala e invocarlas en Spark SQL.

Agregador

Sintaxis Aggregator[-IN, BUF, OUT]

Clase base para agregaciones definidas por el usuario, que se puede usar en operaciones de conjunto de datos para tomar todos los elementos de un grupo y reducirlos a un solo valor.

• IN: tipo de entrada para la agregación.

• BUF: tipo del valor intermedio de la reducción.

• OUT: tipo del resultado final de la salida.

• bufferEncoder: Encoder[BUF]

  Codificador del tipo de valor intermedio.

• finish(reduction: BUF): OUT

  Transforme la salida de la reducción.

• merge(b1: BUF, b2: BUF): BUF

  Combine dos valores intermedios.

• outputEncoder: Encoder[OUT]

  Codificador del tipo de valor de salida final.

• reduce(b: BUF, a: IN): BUF

  Agregue el valor de entrada a en el valor intermedio actual. Para mejorar el rendimiento, la función puede modificar b y devolverlo en lugar de construir un nuevo objeto para b.

• zero: BUF

  Valor inicial del resultado intermedio para esta agregación.

Ejemplos

Funciones de agregado definidas por el usuario con seguridad de tipos

Las agregaciones definidas por el usuario para conjuntos de datos fuertemente tipados giran en torno a la clase abstracta Aggregator. Por ejemplo, un promedio definido por el usuario con seguridad de tipos puede tener el siguiente aspecto:

Scala

import org.apache.spark.sql.{Encoder, Encoders, SparkSession}
import org.apache.spark.sql.expressions.Aggregator

case class Employee(name: String, salary: Long)
case class Average(var sum: Long, var count: Long)

object MyAverage extends Aggregator[Employee, Average, Double] {
  // A zero value for this aggregation. Should satisfy the property that any b + zero = b
  def zero: Average = Average(0L, 0L)
  // Combine two values to produce a new value. For performance, the function may modify `buffer`
  // and return it instead of constructing a new object
  def reduce(buffer: Average, employee: Employee): Average = {
    buffer.sum += employee.salary
    buffer.count += 1
    buffer
  }
  // Merge two intermediate values
  def merge(b1: Average, b2: Average): Average = {
    b1.sum += b2.sum
    b1.count += b2.count
    b1
  }
  // Transform the output of the reduction
  def finish(reduction: Average): Double = reduction.sum.toDouble / reduction.count
  // The Encoder for the intermediate value type
  val bufferEncoder: Encoder[Average] = Encoders.product
  // The Encoder for the final output value type
  val outputEncoder: Encoder[Double] = Encoders.scalaDouble
}

Java

import java.io.Serializable;

import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoder;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SparkSession;
import org.apache.spark.sql.TypedColumn;
import org.apache.spark.sql.expressions.Aggregator;

public static class Employee implements Serializable {
    private String name;
    private long salary;

    // Constructors, getters, setters...
}

public static class Average implements Serializable  {
  private long sum;
  private long count;

  // Constructors, getters, setters...
}

public static class MyAverage extends Aggregator<Employee, Average, Double> {
  // A zero value for this aggregation. Should satisfy the property that any b + zero = b
  public Average zero() {
    return new Average(0L, 0L);
  }
  // Combine two values to produce a new value. For performance, the function may modify `buffer`
  // and return it instead of constructing a new object
  public Average reduce(Average buffer, Employee employee) {
    long newSum = buffer.getSum() + employee.getSalary();
    long newCount = buffer.getCount() + 1;
    buffer.setSum(newSum);
    buffer.setCount(newCount);
    return buffer;
  }
  // Merge two intermediate values
  public Average merge(Average b1, Average b2) {
    long mergedSum = b1.getSum() + b2.getSum();
    long mergedCount = b1.getCount() + b2.getCount();
    b1.setSum(mergedSum);
    b1.setCount(mergedCount);
    return b1;
  }
  // Transform the output of the reduction
  public Double finish(Average reduction) {
    return ((double) reduction.getSum()) / reduction.getCount();
  }
  // The Encoder for the intermediate value type
  public Encoder<Average> bufferEncoder() {
    return Encoders.bean(Average.class);
  }
  // The Encoder for the final output value type
  public Encoder<Double> outputEncoder() {
    return Encoders.DOUBLE();
  }
}

Encoder<Employee> employeeEncoder = Encoders.bean(Employee.class);
String path = "examples/src/main/resources/employees.json";
Dataset<Employee> ds = spark.read().format("json").load(path).as(employeeEncoder);
ds.show();
// +-------+------+
// |   name|salary|
// +-------+------+
// |Michael|  3000|
// |   Andy|  4500|
// | Justin|  3500|
// |  Berta|  4000|
// +-------+------+

MyAverage myAverage = new MyAverage();
// Convert the function to a `TypedColumn` and give it a name
TypedColumn<Employee, Double> averageSalary = myAverage.toColumn().name("average_salary");
Dataset<Double> result = ds.select(averageSalary);
result.show();
// +--------------+
// |average_salary|
// +--------------+
// |        3750.0|
// +--------------+

Funciones de agregado definidas por el usuario sin tipo

Las agregaciones con tipo, como se ha descrito anteriormente, también se pueden registrar como funciones de agregado definidas por el usuario sin tipo para su uso con DataFrames. Por ejemplo, un promedio definido por el usuario para DataFrames sin tipo puede tener el siguiente aspecto:

Scala

import org.apache.spark.sql.{Encoder, Encoders, SparkSession}
import org.apache.spark.sql.expressions.Aggregator
import org.apache.spark.sql.functions

case class Average(var sum: Long, var count: Long)

object MyAverage extends Aggregator[Long, Average, Double] {
  // A zero value for this aggregation. Should satisfy the property that any b + zero = b
  def zero: Average = Average(0L, 0L)
  // Combine two values to produce a new value. For performance, the function may modify `buffer`
  // and return it instead of constructing a new object
  def reduce(buffer: Average, data: Long): Average = {
    buffer.sum += data
    buffer.count += 1
    buffer
  }
  // Merge two intermediate values
  def merge(b1: Average, b2: Average): Average = {
    b1.sum += b2.sum
    b1.count += b2.count
    b1
  }
  // Transform the output of the reduction
  def finish(reduction: Average): Double = reduction.sum.toDouble / reduction.count
  // The Encoder for the intermediate value type
  val bufferEncoder: Encoder[Average] = Encoders.product
  // The Encoder for the final output value type
  val outputEncoder: Encoder[Double] = Encoders.scalaDouble
}

// Register the function to access it
spark.udf.register("myAverage", functions.udaf(MyAverage))

val df = spark.read.format("json").load("examples/src/main/resources/employees.json")
df.createOrReplaceTempView("employees")
df.show()
// +-------+------+
// |   name|salary|
// +-------+------+
// |Michael|  3000|
// |   Andy|  4500|
// | Justin|  3500|
// |  Berta|  4000|
// +-------+------+

val result = spark.sql("SELECT myAverage(salary) as average_salary FROM employees")
result.show()
// +--------------+
// |average_salary|
// +--------------+
// |        3750.0|
// +--------------+

Java

import java.io.Serializable;

import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoder;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;
import org.apache.spark.sql.expressions.Aggregator;
import org.apache.spark.sql.functions;

public static class Average implements Serializable  {
    private long sum;
    private long count;

    // Constructors, getters, setters...

}

public static class MyAverage extends Aggregator<Long, Average, Double> {
  // A zero value for this aggregation. Should satisfy the property that any b + zero = b
  public Average zero() {
    return new Average(0L, 0L);
  }
  // Combine two values to produce a new value. For performance, the function may modify `buffer`
  // and return it instead of constructing a new object
  public Average reduce(Average buffer, Long data) {
    long newSum = buffer.getSum() + data;
    long newCount = buffer.getCount() + 1;
    buffer.setSum(newSum);
    buffer.setCount(newCount);
    return buffer;
  }
  // Merge two intermediate values
  public Average merge(Average b1, Average b2) {
    long mergedSum = b1.getSum() + b2.getSum();
    long mergedCount = b1.getCount() + b2.getCount();
    b1.setSum(mergedSum);
    b1.setCount(mergedCount);
    return b1;
  }
  // Transform the output of the reduction
  public Double finish(Average reduction) {
    return ((double) reduction.getSum()) / reduction.getCount();
  }
  // The Encoder for the intermediate value type
  public Encoder<Average> bufferEncoder() {
    return Encoders.bean(Average.class);
  }
  // The Encoder for the final output value type
  public Encoder<Double> outputEncoder() {
    return Encoders.DOUBLE();
  }
}

// Register the function to access it
spark.udf().register("myAverage", functions.udaf(new MyAverage(), Encoders.LONG()));

Dataset<Row> df = spark.read().format("json").load("examples/src/main/resources/employees.json");
df.createOrReplaceTempView("employees");
df.show();
// +-------+------+
// |   name|salary|
// +-------+------+
// |Michael|  3000|
// |   Andy|  4500|
// | Justin|  3500|
// |  Berta|  4000|
// +-------+------+

Dataset<Row> result = spark.sql("SELECT myAverage(salary) as average_salary FROM employees");
result.show();
// +--------------+
// |average_salary|
// +--------------+
// |        3750.0|
// +--------------+

SQL

-- Compile and place UDAF MyAverage in a JAR file called `MyAverage.jar` in /tmp.
CREATE FUNCTION myAverage AS 'MyAverage' USING JAR '/tmp/MyAverage.jar';

SHOW USER FUNCTIONS;
+------------------+
|          function|
+------------------+
| default.myAverage|
+------------------+

CREATE TEMPORARY VIEW employees
USING org.apache.spark.sql.json
OPTIONS (
    path "examples/src/main/resources/employees.json"
);

SELECT * FROM employees;
+-------+------+
|   name|salary|
+-------+------+
|Michael|  3000|
|   Andy|  4500|
| Justin|  3500|
|  Berta|  4000|
+-------+------+

SELECT myAverage(salary) as average_salary FROM employees;
+--------------+
|average_salary|
+--------------+
|        3750.0|
+--------------+

Instrucciones relacionadas

• Funciones escalares definidas por el usuario (UDF)
• Integración con las UDF, UDAF y UDTF de Hive