Documentation

Reduction
Trail: Collections
Lesson: Aggregate Operations

Reduction

The section Aggregate Operations describes the following pipeline of operations, which calculates the average age of all male members in the collection roster:

double average = roster
    .stream()
    .filter(p -> p.getGender() == Person.Sex.MALE)
    .mapToInt(Person::getAge)
    .average()
    .getAsDouble();

The JDK contains many terminal operations (such as average, sum, min, max, and count) that return one value by combining the contents of a stream. These operations are called reduction operations. The JDK also contains reduction operations that return a collection instead of a single value. Many reduction operations perform a specific task, such as finding the average of values or grouping elements into categories. However, the JDK provides you with the general-purpose reduction operations reduce and collect, which this section describes in detail.

This section covers the following topics:

You can find the code excerpts described in this section in the example ReductionExamples.

The Stream.reduce Method

The Stream.reduce method is a general-purpose reduction operation. Consider the following pipeline, which calculates the sum of the male members' ages in the collection roster. It uses the Stream.sum reduction operation:

Integer totalAge = roster
    .stream()
    .mapToInt(Person::getAge)
    .sum();

Compare this with the following pipeline, which uses the Stream.reduce operation to calculate the same value:

Integer totalAgeReduce = roster
   .stream()
   .map(Person::getAge)
   .reduce(
       0,
       (a, b) -> a + b);

The reduce operation in this example takes two arguments:

The reduce operation always returns a new value. However, the accumulator function also returns a new value every time it processes an element of a stream. Suppose that you want to reduce the elements of a stream to a more complex object, such as a collection. This might hinder the performance of your application. If your reduce operation involves adding elements to a collection, then every time your accumulator function processes an element, it creates a new collection that includes the element, which is inefficient. It would be more efficient for you to update an existing collection instead. You can do this with the Stream.collect method, which the next section describes.

The Stream.collect Method

Unlike the reduce method, which always creates a new value when it processes an element, the collect method modifies, or mutates, an existing value.

Consider how to find the average of values in a stream. You require two pieces of data: the total number of values and the sum of those values. However, like the reduce method and all other reduction methods, the collect method returns only one value. You can create a new data type that contains member variables that keep track of the total number of values and the sum of those values, such as the following class, Averager:

class Averager implements IntConsumer
{
    private int total = 0;
    private int count = 0;
        
    public double average() {
        return count > 0 ? ((double) total)/count : 0;
    }
        
    public void accept(int i) { total += i; count++; }
    public void combine(Averager other) {
        total += other.total;
        count += other.count;
    }
}

The following pipeline uses the Averager class and the collect method to calculate the average age of all male members:

Averager averageCollect = roster.stream()
    .filter(p -> p.getGender() == Person.Sex.MALE)
    .map(Person::getAge)
    .collect(Averager::new, Averager::accept, Averager::combine);
                   
System.out.println("Average age of male members: " +
    averageCollect.average());

The collect operation in this example takes three arguments:

Note the following:

Although the JDK provides you with the average operation to calculate the average value of elements in a stream, you can use the collect operation and a custom class if you need to calculate several values from the elements of a stream.

The collect operation is best suited for collections. The following example puts the names of the male members in a collection with the collect operation:

List<String> namesOfMaleMembersCollect = roster
    .stream()
    .filter(p -> p.getGender() == Person.Sex.MALE)
    .map(p -> p.getName())
    .collect(Collectors.toList());

This version of the collect operation takes one parameter of type Collector. This class encapsulates the functions used as arguments in the collect operation that requires three arguments (supplier, accumulator, and combiner functions).

The Collectors class contains many useful reduction operations, such as accumulating elements into collections and summarizing elements according to various criteria. These reduction operations return instances of the class Collector, so you can use them as a parameter for the collect operation.

This example uses the Collectors.toList operation, which accumulates the stream elements into a new instance of List. As with most operations in the Collectors class, the toList operator returns an instance of Collector, not a collection.

The following example groups members of the collection roster by gender:

Map<Person.Sex, List<Person>> byGender =
    roster
        .stream()
        .collect(
            Collectors.groupingBy(Person::getGender));

The groupingBy operation returns a map whose keys are the values that result from applying the lambda expression specified as its parameter (which is called a classification function). In this example, the returned map contains two keys, Person.Sex.MALE and Person.Sex.FEMALE. The keys' corresponding values are instances of List that contain the stream elements that, when processed by the classification function, correspond to the key value. For example, the value that corresponds to key Person.Sex.MALE is an instance of List that contains all male members.

The following example retrieves the names of each member in the collection roster and groups them by gender:

Map<Person.Sex, List<String>> namesByGender =
    roster
        .stream()
        .collect(
            Collectors.groupingBy(
                Person::getGender,                      
                Collectors.mapping(
                    Person::getName,
                    Collectors.toList())));

The groupingBy operation in this example takes two parameters, a classification function and an instance of Collector. The Collector parameter is called a downstream collector. This is a collector that the Java runtime applies to the results of another collector. Consequently, this groupingBy operation enables you to apply a collect method to the List values created by the groupingBy operator. This example applies the collector mapping, which applies the mapping function Person::getName to each element of the stream. Consequently, the resulting stream consists of only the names of members. A pipeline that contains one or more downstream collectors, like this example, is called a multilevel reduction.

The following example retrieves the total age of members of each gender:

Map<Person.Sex, Integer> totalAgeByGender =
    roster
        .stream()
        .collect(
            Collectors.groupingBy(
                Person::getGender,                      
                Collectors.reducing(
                    0,
                    Person::getAge,
                    Integer::sum)));

The reducing operation takes three parameters:

The following example retrieves the average age of members of each gender:

Map<Person.Sex, Double> averageAgeByGender = roster
    .stream()
    .collect(
        Collectors.groupingBy(
            Person::getGender,                      
            Collectors.averagingInt(Person::getAge)));

Previous page: Aggregate Operations
Next page: Parallelism