The document discusses stream processing in Java, including key concepts like streams, mappers, reducers, and collectors. It explains that streams provide a way to process sequences of elements using aggregate operations. Streams are lazy and allow chaining intermediate operations like mapping and filtering to transform data before a terminal operation like counting or summing. Common stream operations include mapping elements to new forms, filtering values, and reducing elements to single values through aggregation.
The document discusses stream processing in Java, including key concepts like streams, mappers, reducers, and collectors. It explains that streams provide a way to process sequences of elements using aggregate operations. Streams are lazy and allow chaining intermediate operations like mapping and filtering to transform data before a terminal operation like counting or summing. Common stream operations include mapping elements to new forms, filtering values, and reducing elements to single values through aggregation.
The document discusses stream processing in Java, including key concepts like streams, mappers, reducers, and collectors. It explains that streams provide a way to process sequences of elements using aggregate operations. Streams are lazy and allow chaining intermediate operations like mapping and filtering to transform data before a terminal operation like counting or summing. Common stream operations include mapping elements to new forms, filtering values, and reducing elements to single values through aggregation.
The document discusses stream processing in Java, including key concepts like streams, mappers, reducers, and collectors. It explains that streams provide a way to process sequences of elements using aggregate operations. Streams are lazy and allow chaining intermediate operations like mapping and filtering to transform data before a terminal operation like counting or summing. Common stream operations include mapping elements to new forms, filtering values, and reducing elements to single values through aggregation.
• Computation on the source data is only performed when the terminal operation is initiated.
• Source elements are consumed only as needed.
• Preferably, one element passes the complete pipeline before the next is polled. • Enhanced compute performance may be observed (due to CPU-cache affinity).
• Memory footprint of most provided stream operations is O(1)
• ensuring that large data sources can be processed with low memory footprint.
3 07/07/2023
METHOD-CHAINING IN STREAMS
• Stream operations can be ‘method-chained’: int numPassed = grades● stream()
(a.k.a. ‘dot-chaining’) ● filter( x à x >= 5 ) ● count();
• This contributes to • flexibility of applying combinations of operations in any order • brevity of code • readability of code
See also: https://en.wikipedia.org/wiki/Method_chaining
• parallelStream() will try to leverage multi-threading
on multiple processor cores to speed-up execution • This requires all code to be ‘thread-safe’ • This is a topic for courses on ‘Parallel Computing’ (do not use parallel streams for now…)
reduce( Optional<T> reduction starts empty BinaryOperator<T> accumulator) reduce(T identity, T uses identity as the start value BinaryOperator<T> accumulator) (e.g. identity=0 for sum reduction) reduce(U identity, U integrates mapping and accumulation BiFunction<U, ? super T, U> accumulator from type T to type U BinaryOperator<T> combiner
15
15
EXAMPLE
• Find the book with the longest title:
BinaryOperator<Book> which each time retains the book get the final surviving book with the longest title from Optional<Book>
